A method for identifying dual-diode models of photovoltaic cells based on improved bat algorithm

Abstract

The present invention discloses a photovoltaic cell dual-diode model identification method based on an improved bat algorithm, firstly tests and obtains the I-V curve of the tested photovoltaic cell under the standard test conditions of the photovoltaic cell, and then uses the improved bat algorithm to identify seven photovoltaic cell dual-diode model parameters of the tested photovoltaic cell, including: step S2-1, determining the parameter value range; step S2-2, initializing the bat population; step S2-3, adjusting the bat population position; step S2-4, calculating the fitness of the population individuals; step S2-5, premature convergence control update strategy; step S2-6, judging whether the bat position is out of bounds; step S2-7, updating the optimal bat position; step S2-8, when the number of iterations reaches the maximum, outputting the optimal bat position. The improved bat algorithm of the present invention is improved in three aspects: initial position, premature convergence control and out-of-bounds check, has the advantages of high accuracy, and is conducive to photovoltaic cell simulation calculation.

Description

Technical Field

[0001] The invention relates to a model parameter identification method of a photovoltaic cell using a photovoltaic cell double-diode model, in particular to a photovoltaic cell double-diode model identification method based on an improved bat algorithm. Background Art

[0002] In order to cope with the pollution of fossil energy and the impact of the greenhouse effect, new energy, especially photovoltaic power generation systems, has attracted extensive research interest. As an important component of the photovoltaic system, the operating conditions of photovoltaic modules vary due to various conditions such as battery type, weather conditions, aging conditions, etc. Although the impact of these conditions on the operating conditions can be measured through experimental or field empirical methods, this method has the disadvantages of high investment and inconvenient adjustment, and the modeling and simulation method can make up for this disadvantage well.

[0003] There are many models of photovoltaic cells, the most commonly used are the single diode model and the double diode model. Compared with the single diode model, the double diode model has one more diode and performs better in low irradiation environments. The double diode model contains seven parameters to be estimated, which is a large number of parameters and difficult to identify. The existing technology often uses meta-heuristic algorithms. Although this method is simple to implement and has a wide range of applications, it has a large amount of calculations and is prone to falling into local optimal values.

[0004] Among them, Figure 1 As shown, it is known that the photovoltaic cell double diode model (DDM) satisfies the port voltage-current function relationship shown in the following formula 1:

[0005]

[0006] In the formula, R s is the series resistance parameter, R sh is the parallel resistance parameter, I is the port output current, I ph is the photocurrent parameter of the battery, I o1 is the reverse saturation current parameter of diode D1 and is recorded as the first reverse saturation current parameter, I o2 is the reverse saturation current parameter of diode D2 and is recorded as the second reverse saturation current parameter, exp is the exponential power of the natural constant e, V is the port output voltage, n1 is the ideal factor parameter of diode D1 and is recorded as the first ideal factor parameter, n2 is the ideal factor parameter of diode D2 and is recorded as the second ideal factor parameter, V t is the temperature potential.

[0007] Determine n1, n2, and R of the photovoltaic cell double diode model (DDM) s ,I ph ,I o1 ,I o2 and Rsh The process of determining the values ​​of these seven parameters is called parameter identification of the photovoltaic cell double diode model. Summary of the invention

[0008] The technical problem to be solved by the present invention is to provide a photovoltaic cell dual diode model identification method based on an improved bat algorithm to solve the problems of large calculation amount and low accuracy in the model parameter identification method of photovoltaic cells using the photovoltaic cell dual diode model in the prior art.

[0009] To solve the above technical problems, the technical solution adopted by the present invention is as follows:

[0010] A photovoltaic cell double diode model identification method based on an improved bat algorithm is applicable to photovoltaic cells under test using a photovoltaic cell double diode model (DDM);

[0011] The invention is characterized by comprising:

[0012] Step S1, testing and obtaining an IV curve of the photovoltaic cell under standard test conditions (STC) of the photovoltaic cell, where I represents the port output current of the photovoltaic cell under test, and V represents the port output voltage of the photovoltaic cell under test;

[0013] Step S2: using an improved bat algorithm to identify seven photovoltaic cell dual diode model parameters of the photovoltaic cell under test.

[0014] Among them, Figure 2 As shown, the identification process of the improved bat algorithm includes:

[0015] Step S2-1, setting the parameter value range of the photovoltaic cell under test, including: maximum position vector parameter x max =[n 1max ,n 2max ,R smax ,I phmax ,I o1max ,I o2m ,R shmax ], minimum position vector parameter x min =[n 1min ,n 2min ,R smin ,I phmin ,I o1min ,I o2min ,R shmin ]、Frequency upper limit f max and the lower frequency limit f min ; where n 1max 、n 2max , R smax ,I phmax ,I o1max ,Io2ma , R shmax They are the first ideal factor parameter upper limit, the second ideal factor parameter upper limit, the series resistance parameter upper limit, the battery photocurrent parameter upper limit, the first reverse saturation current parameter upper limit, the second reverse saturation current parameter upper limit, and the parallel resistance parameter upper limit; n 1min 、n 2min , R smin ,I phmin ,I o1min ,I 02min , R shmin They are, in order, the first ideal factor parameter lower limit value, the second ideal factor parameter lower limit value, the series resistance parameter lower limit value, the battery photocurrent parameter lower limit value, the first reverse saturation current parameter lower limit value, the second reverse saturation current parameter lower limit value, and the parallel resistance parameter lower limit value;

[0016] Step S2-2: When the number of iterations t = 0 (i.e. before starting the iteration), set the maximum number of iterations t max and the number of bats M contained in the bat population, so as to control the computational complexity and computational accuracy of the improved bat algorithm; and the values ​​of the bat information of each bat in the bat population are randomly initialized, and the random values ​​obtained by random initialization are recorded as initialization values, wherein the bat information includes: emission rate, loudness, speed, frequency and position vector parameter x = [n1, n2, R s ,I ph ,I o1 ,I o2 ,R sh ], the initialization values ​​are recorded as: r 0 , A 0 、v 0 、f 0 and n1, n2, R s ,I ph ,I o1 ,I o2 , R sh They are the first ideal factor parameter, the second ideal factor parameter, the series resistance parameter, the cell photocurrent parameter, the first reverse saturation current parameter, the second reverse saturation current parameter, and the parallel resistance parameter in the photovoltaic cell double diode model (DDM);

[0017] Step S2-3, for each bat: use the initialization values ​​of the first four parameters in the IV curve and position vector parameter x According to the following formulas 2 to 4, the last three parameters I in the position vector parameter x are o1 ,I o2 , R sh The value of is used to adjust the position:

[0018]

[0019] In the formula, X1, X2, The first reverse saturation current parameter I o1 , the second reverse saturation current parameter I o2 , Parallel resistance parameter R sh The value after the position adjustment;

[0020] A and b both represent matrices. T represents the transposed matrix of matrix A, (A T A) -1 Represents the matrix A T The inverse matrix of A;

[0021] exp is the exponential power of the natural constant e, N is the number of sampling points of the IV curve, V1,…,V N and I1,…,I N are the port output voltage and port output current from the first sampling point to the Nth sampling point in the IV curve, V t is the temperature potential of 25.7 mV under the standard test conditions (STC) of the photovoltaic cell;

[0022] In addition, one of the M bats is randomly selected as the optimal bat of the 0th iteration. The position vector parameter x of the initial optimal bat calculated by formula 2 is recorded as the optimal position vector parameter of the optimal bat of the 0th iteration:

[0023] After completing step S2-1 to step S2-3, the iteration is started.

[0024] Step S2-4, calculate the fitness F(x) of each bat in the bat population according to the following formula 5:

[0025]

[0026] Where V1,…,V N ,I1,…,I N and V t The meaning of is the same as that in Formula 4; |·| indicates the absolute value; when the number of iterations t = 0, n1, n2, R s ,I ph ,I o1 ,I o2 , R sh The value of the position vector parameter x is calculated using formula 2 in step S2-3 X1, X2, Calculate; when the number of iterations t≥1, n1, n2, Rs ,I ph ,I o1 ,I o2 , R sh The calculation is performed using the value of the position vector parameter x updated after steps S2-5 to S2-7;

[0027] Step S2-5, adopting the premature convergence control strategy to update the position vector parameter x of each bat in the bat population, that is: first, the M1 bats with the smallest fitness F(x) among the M bats update the position vector parameter x according to the local search method; then, randomly select M2 bats from the remaining M-M1 bats, and update the position vector parameter x through chaotic mutation; finally, the remaining M-M1-M2 bats update the position vector parameter x according to the global search method; wherein, the values ​​of the number of local search bats M1 and the number of chaotic mutation bats M2 are determined according to Formula 6:

[0028]

[0029] In the formula, min is the minimum function, ceil is the upward rounding function, and floor is the downward rounding function;

[0030] Step S2-6: for each bat whose position vector parameter x is updated in step S2-5, determine whether the parameter in the updated position vector parameter x is out of bounds, that is, determine whether the value θx′ of the parameter exceeds [x min ,x max ], and the parameters that are judged to be beyond the range are recorded as out-of-bounds parameters; for example: n1 exceeds [n 1min ,n 1max ], or n2 exceeds the range of [n 2min ,n 2max ];

[0031] The values ​​of out-of-bounds parameters are adjusted as follows:

[0032] If θx′ is greater than the maximum position vector parameter x max The corresponding parameter θx in max , then the out-of-bounds parameter is adjusted to:

[0033]

[0034] If θx′ is less than the minimum position vector parameter x min The corresponding parameter θx in min , then the out-of-bounds parameter is adjusted to:

[0035]

[0036] Wherein, θx represents the value after the out-of-bounds parameter is adjusted, and θx′ represents the value before the out-of-bounds parameter is adjusted, that is, the parameter value corresponding to θx in the position vector parameter x updated in step S2-5; θx max and θx min Respectively represent the maximum position vector parameter x max and the minimum position vector parameter x min The parameter corresponding to θx′ in ; k1 and k2 are the first auxiliary judgment parameter and the second auxiliary judgment parameter respectively; mod is the modulus operator, k1mod 2=0 means that the remainder of the division of k1 and 2 is 0;

[0037] Step S2-7, update the optimal bat in the following manner:

[0038] First, a random number β1 is generated within a range of values ​​adapted to the emission rate, and a random number β2 is generated within a range of values ​​adapted to the loudness; and the emission rate of the optimal bat at the t+1th iteration is and loudness Update according to Formula 7 and Formula 8;

[0039] Then, determine whether

[0040] If the result of the judgment is yes, then: according to Formula 2 to Formula 4, the position vector parameter x of each bat in the tth iteration adjusted by step S2-6 is adjusted, wherein the first four parameters in the position vector parameter x are regarded as the initialization values ​​described in step S2-3 And, according to the position vector parameter x adjusted in this step, the fitness F(x) of each bat is calculated according to the formula 5, and the bat with the smallest fitness F(x) and its position vector parameter x are recorded as the new optimal bat and its position vector parameter The firing rate and loudness of the new optimal bat are respectively and

[0041] If the judgment result is no, then: directly go to step S2-8;

[0042] Finally, determine whether and in, represents the fitness of the new optimal bat, represents the fitness of the best bat in the tth iteration;

[0043] If the result is yes, then: the optimal bat and its position vector parameters of the t+1th iteration are Update to the new optimal bat and its position vector parameters The firing rate of the optimal bat in the t+1th iteration is and loudness Updated to and And, go to step S2-8;

[0044] If the judgment result is no, then: the optimal bat of the t+1th iteration is maintained as the optimal bat of the tth iteration, and the process goes to step S2-8;

[0045] Step S2-8, the number of iterations t is increased by 1; and when t <t max When , go to step 2-4 to repeat steps 2-4 to 2-8; otherwise, output the optimal position vector parameter of the optimal bat of the tth iteration As the seven photovoltaic cell double diode model parameters of the photovoltaic cell under test.

[0046] Wherein, in the step S2-5, the steps of the local search method are: updating the bat's position vector parameter x according to the following formulas 7 to 9;

[0047]

[0048] In the formula, is the firing rate of the i-th bat at the t+1th iteration, is the initialization value of the firing rate of the i-th bat, and γ is the firing rate enhancement coefficient between 0 and 1;

[0049] is the loudness of the i-th bat at the t+1th iteration, is the loudness initialization value of the i-th bat, and α is the loudness attenuation coefficient between 0 and 1;

[0050] is the updated position vector parameter of the i-th bat at the t+1th iteration, is the position vector parameter of the i-th bat calculated by formula 2 in step S2-3, is the random number of the i-th bat at the t-th iteration, t+1 > is the average loudness of all M bats at the t+1th iteration.

[0051] Among them, in step S2-5, the chaotic mutation step is:

[0052] First, the position vector parameter of the i-th bat is updated at the t-th iteration according to formula 10: Normalize:

[0053] ​

[0054] In the formula, represents the normalized position vector parameter of the i-th bat at the t-th iteration;

[0055] Then, according to formula 11, the normalized position vector parameters are calculated Piecewise linear chaotic map (PWLCM):

[0056]

[0057] In the formula, represents the mapped position vector parameter of the i-th bat at the t-th iteration; p is the control parameter, and p∈(0,0.5), that is, the chaotic state, and the optimal value of p is 0.499;

[0058] Finally, according to Formula 12, the mapped position vector parameter Restored to the position vector parameter updated by the i-th bat at the t+1th iteration To update the bat's position vector parameter x:

[0059]

[0060] Among them, in step S2-5, the steps of the global search method are: according to formula 13 to formula 15, update the bat's position vector parameter x;

[0061] f i =f min +(f max -f min )β i [Formula 13]

[0062]

[0063] In the formula, f i represents the sound wave frequency of the i-th bat, f max and f min are the upper and lower frequency limits, respectively, i is the random number of the i-th bat; is the speed of the i-th bat at the t+1th iteration, Initialize the speed of the i-th bat, is the optimal position vector parameter at the tth iteration, is the optimal position vector parameter of the optimal bat selected in step S2-3, is the updated position vector parameter of the i-th bat at the t+1th iteration, is the position vector parameter of the i-th bat calculated by formula 2 in step S2-3.

[0064] Compared with the prior art, the present invention has the following beneficial effects:

[0065] The present invention adopts an improved bat algorithm to identify seven photovoltaic cell double diode model parameters of the photovoltaic cell under test, which has the advantage of high accuracy and is beneficial to photovoltaic cell simulation calculation.

[0066] Specifically, the improved bat algorithm of the present invention is improved in three aspects: initial position, premature convergence control and out-of-bounds check.

[0067] Initial position: The present invention adopts a partially randomized method to perform parameter initialization assignment before iteration, that is, firstly, all parameters in the position vector parameter x are randomly initialized through step 2-2 to obtain initialization values ​​that are all randomly taken, and then, through step 2-3, the initialization values ​​of the first four parameters in the position vector parameter x and the IV curve are used to calculate and adjust the values ​​of the last three parameters in the position vector parameter x, so that the adjusted position vector parameter x obtained in step 2-3 has the first ideal factor parameter n1, the second ideal factor parameter n2, and the series resistance parameter R which are randomly taken. s 、Battery photocurrent parameter I ph , and the first reverse saturation current parameter I o1 , the second reverse saturation current parameter I o2 , Parallel resistance parameter R sh Therefore, compared with the completely random initialization, the present invention has higher credibility and helps to improve the recognition accuracy.

[0068] In terms of premature convergence control: step S2-5 of the present invention enables the bat population to have the capabilities of local search, chaotic mutation and global search at the same time, and controls the number of bats corresponding to the three capabilities, namely: the M1 bats with the smallest fitness F(x) update the position vector parameter x according to the local search method, and randomly select M2 bats from the remaining M-M1 bats to update the position vector parameter x through chaotic mutation; the remaining M-M1-M2 bats update the position vector parameter x according to the global search method; so that there are more bats in the early chaotic mutation and global search, which is conducive to the traversal of the solution space, and more bats in the middle and late search local search, which is conducive to the convergence of the algorithm. Therefore, the search accuracy of the bat population is improved, the problem of premature convergence of the algorithm is avoided to a certain extent, and it is helpful to improve the recognition accuracy.

[0069] In terms of out-of-bounds checking: the present invention abandons the method of setting the speed of out-of-bounds bats to 0 adopted in the prior art, and remaps the out-of-bounds parameters of the bats to the search space through step 2-1, thereby retaining the optimization ability of the bats and helping to improve the recognition accuracy. BRIEF DESCRIPTION OF THE DRAWINGS

[0070] The present invention is further described in detail below in conjunction with the accompanying drawings and specific embodiments:

[0071] Figure 1 It is the equivalent circuit diagram of the photovoltaic cell double diode model (DDM);

[0072] Figure 2 It is a flowchart of the identification process of the improved bat algorithm in the present invention;

[0073] Figure 3 The JV curve simulation comparison results of the present invention and the prior art in the comparative experiment are shown in FIG. DETAILED DESCRIPTION

[0074] The present invention is described in detail below in conjunction with the embodiments and the accompanying drawings to help those skilled in the art better understand the inventive concept of the present invention. However, the protection scope of the claims of the present invention is not limited to the following embodiments. For those skilled in the art, all other embodiments obtained without creative work without departing from the inventive concept of the present invention belong to the protection scope of the present invention.

[0075] like Figure 2 As shown, the present invention discloses a photovoltaic cell double diode model identification method based on an improved bat algorithm, which is applicable to a photovoltaic cell under test using a photovoltaic cell double diode model (DDM);

[0076] include:

[0077] Step S1, testing and obtaining an IV curve of the photovoltaic cell under standard test conditions (STC) of the photovoltaic cell, where I represents the port output current of the photovoltaic cell under test, and V represents the port output voltage of the photovoltaic cell under test;

[0078] Step S2: using an improved bat algorithm to identify seven photovoltaic cell dual diode model parameters of the photovoltaic cell under test.

[0079] Among them, Figure 2 As shown, the identification process of the improved bat algorithm includes:

[0080] Step S2-1, setting the parameter value range of the photovoltaic cell under test, including: maximum position vector parameter x max =[n 1max ,n 2max ,R smax ,I phmax ,I o1max ,I o2max ,R shmax ], minimum position vector parameter x min =[n 1min ,n 2min ,Rsmin ,I phmin ,I o1mi ,I o2min ,R shmin ]、Frequency upper limit f max and the lower frequency limit f min ; where n imax 、n 2max , R smax ,I phmax ,I oimax ,I o2m , R shmax They are the first ideal factor parameter upper limit, the second ideal factor parameter upper limit, the series resistance parameter upper limit, the battery photocurrent parameter upper limit, the first reverse saturation current parameter upper limit, the second reverse saturation current parameter upper limit, and the parallel resistance parameter upper limit; n 1min 、n 2min , R smin ,I phmin ,I o1min ,I o2min , R shmin They are, in order, the first ideal factor parameter lower limit value, the second ideal factor parameter lower limit value, the series resistance parameter lower limit value, the battery photocurrent parameter lower limit value, the first reverse saturation current parameter lower limit value, the second reverse saturation current parameter lower limit value, and the parallel resistance parameter lower limit value;

[0081] Step S2-2: When the number of iterations t = 0 (i.e. before starting the iteration), set the maximum number of iterations t max and the number of bats M contained in the bat population, so as to control the computational complexity and computational accuracy of the improved bat algorithm; and the values ​​of the bat information of each bat in the bat population are randomly initialized, and the random values ​​obtained by random initialization are recorded as initialization values, wherein the bat information includes: emission rate, loudness, speed, frequency and position vector parameter x = [n1, n2, R s ,I ph ,I o1 ,I o2 ,R sh ], the initialization values ​​are recorded as: r 0 , A 0 、v 0 、f 0 and n1, n2, R s ,I ph ,I o1 ,I o2 , R shThey are the first ideal factor parameter, the second ideal factor parameter, the series resistance parameter, the cell photocurrent parameter, the first reverse saturation current parameter, the second reverse saturation current parameter, and the parallel resistance parameter in the photovoltaic cell double diode model (DDM);

[0082] Step S2-3, for each bat: use the initialization values ​​of the first four parameters in the IV curve and position vector parameter x According to the following formulas 2 to 4, the last three parameters I in the position vector parameter x are o1 ,I o2 , R sh The value of is used to adjust the position:

[0083]

[0084] In the formula, X1, X2, The first reverse saturation current parameter I o1 , the second reverse saturation current parameter I o2 , Parallel resistance parameter R sh The value after the position adjustment;

[0085] A and b both represent matrices. T represents the transposed matrix of matrix A, (A T A) -1 Represents the matrix A T The inverse matrix of A;

[0086] exp is the exponential power of the natural constant e, N is the number of sampling points of the IV curve, V1,…,V N and I1,…,I N are the port output voltage and port output current from the first sampling point to the Nth sampling point in the IV curve, V t is the temperature potential of 25.7 mV under the standard test conditions (STC) of the photovoltaic cell;

[0087] In addition, one of the M bats is randomly selected as the optimal bat of the 0th iteration. The position vector parameter x of the initial optimal bat calculated by formula 2 is recorded as the optimal position vector parameter of the optimal bat of the 0th iteration:

[0088] After completing step S2-1 to step S2-3, the iteration is started.

[0089] Step S2-4, calculate the fitness F(x) of each bat in the bat population according to the following formula 5:

[0090]

[0091] Where V1,…,V N ,I1,…,I N and V t The meaning of is the same as that in Formula 4; |·| indicates the absolute value; when the number of iterations t = 0, n1, n2, R s ,I ph ,I o1 ,I o2 , R sh The value of the position vector parameter x is calculated using formula 2 in step S2-3 X1, X2, Calculate; when the number of iterations t≥1, n1, n2, R s ,I ph ,I o1 ,I o2 , R sh The calculation is performed using the value of the position vector parameter x updated after steps S2-5 to S2-7;

[0092] Step S2-5, adopting the premature convergence control strategy to update the position vector parameter x of each bat in the bat population, that is: first, the M1 bats with the smallest fitness F(x) among the M bats update the position vector parameter x according to the local search method; then, randomly select M2 bats from the remaining M-M1 bats, and update the position vector parameter x through chaotic mutation; finally, the remaining M-M1-M2 bats update the position vector parameter x according to the global search method; wherein, the values ​​of the number of local search bats M1 and the number of chaotic mutation bats M2 are determined according to Formula 6:

[0093]

[0094] In the formula, min is the minimum function, ceil is the upward rounding function, and floor is the downward rounding function;

[0095] Wherein, in the step S2-5, the steps of the local search method are: updating the bat's position vector parameter x according to the following formulas 7 to 9;

[0096]

[0097] In the formula, is the firing rate of the i-th bat at the t+1th iteration, is the initialization value of the firing rate of the i-th bat, and γ is the firing rate enhancement coefficient between 0 and 1;

[0098] is the loudness of the i-th bat at the t+1th iteration, is the loudness initialization value of the i-th bat, and α is the loudness attenuation coefficient between 0 and 1;

[0099] is the updated position vector parameter of the i-th bat at the t+1th iteration, is the position vector parameter of the i-th bat calculated by formula 2 in step S2-3, is the random number of the i-th bat at the t-th iteration, t+1 > is the average loudness of all M bats at the t+1th iteration.

[0100] Among them, in step S2-5, the chaotic mutation step is:

[0101] First, the position vector parameter of the i-th bat is updated at the t-th iteration according to formula 10: Normalize:

[0102]

[0103] In the formula, represents the normalized position vector parameter of the i-th bat at the t-th iteration;

[0104] Then, according to formula 11, the normalized position vector parameters are calculated Piecewise linear chaotic map (PWLCM):

[0105]

[0106] In the formula, represents the mapped position vector parameter of the i-th bat at the t-th iteration; p is the control parameter, and p∈(0,0.5), that is, the chaotic state, and the optimal value of p is 0.499;

[0107] Finally, according to Formula 12, the mapped position vector parameter Restored to the position vector parameter updated by the i-th bat at the t+1th iteration To update the bat's position vector parameter x:

[0108]

[0109] Wherein, in step S2-5, the steps of the global search method are: updating the bat's position vector parameter x according to formulas 13 to 15;

[0110] f i =f min +(f max -f min )β i [Formula 13]​

[0111]

[0112] In the formula, f i represents the sound wave frequency of the i-th bat, f max and f min are the upper and lower frequency limits, respectively, i is the random number of the i-th bat; is the speed of the i-th bat at the t+1th iteration, Initialize the speed of the i-th bat, is the optimal position vector parameter at the tth iteration, is the optimal position vector parameter of the optimal bat selected in step S2-3, is the updated position vector parameter of the i-th bat at the t+1th iteration, is the position vector parameter of the i-th bat calculated by formula 2 in step S2-3.

[0113] Step S2-6: for each bat whose position vector parameter x is updated in step S2-5, determine whether the parameter in the updated position vector parameter x is out of bounds, that is, determine whether the value θx′ of the parameter exceeds [x min ,x max ], and the parameters that are judged to be beyond the range are recorded as out-of-bounds parameters; for example: n1 exceeds [n 1min ,n 1max ], or n2 exceeds the range of [n 2min ,n 2max ];

[0114] The values ​​of out-of-bounds parameters are adjusted as follows:

[0115] If θx′ is greater than the maximum position vector parameter x max The corresponding parameter θx in max , then the out-of-bounds parameter is adjusted to:

[0116]

[0117]

[0118] If θx′ is less than the minimum position vector parameter x min The corresponding parameter θx in min , then the out-of-bounds parameter is adjusted to:

[0119]

[0120] Wherein, θx represents the value after the out-of-bounds parameter is adjusted, and θx′ represents the value before the out-of-bounds parameter is adjusted, that is, the parameter value corresponding to θx in the position vector parameter x updated in step S2-5; θx max and θx min Respectively represent the maximum position vector parameter x max and the minimum position vector parameter x min The parameter corresponding to θx′ in ; k1 and k2 are the first auxiliary judgment parameter and the second auxiliary judgment parameter respectively; mod is the modulus operator, k1mod 2=0 means that the remainder of the division of k1 and 2 is 0;

[0121] Step S2-7, update the optimal bat in the following manner:

[0122] First, a random number β1 is generated within a range of values ​​adapted to the emission rate, and a random number β2 is generated within a range of values ​​adapted to the loudness; and the emission rate of the optimal bat at the t+1th iteration is and loudness Update according to Formula 7 and Formula 8;

[0123] Then, determine whether

[0124] If the result of the judgment is yes, then: according to Formula 2 to Formula 4, the position vector parameter x of each bat in the tth iteration adjusted by step S2-6 is adjusted, wherein the first four parameters in the position vector parameter x are regarded as the initialization values ​​described in step S2-3 And, according to the position vector parameter x adjusted in this step, the fitness F(x) of each bat is calculated according to the formula 5, and the bat with the smallest fitness F(x) and its position vector parameter x are recorded as the new optimal bat and its position vector parameter The firing rate and loudness of the new optimal bat are respectively and

[0125] If the judgment result is no, then: directly go to step S2-8;

[0126] Finally, determine whether and in, represents the fitness of the new optimal bat, represents the fitness of the best bat in the tth iteration;

[0127] If the result is yes, then: the optimal bat and its position vector parameters of the t+1th iteration are Update to the new optimal bat and its position vector parameters The firing rate of the optimal bat in the t+1th iteration is and loudness Updated to and And, go to step S2-8;

[0128] If the judgment result is no, then: the optimal bat of the t+1th iteration is maintained as the optimal bat of the tth iteration, and the process goes to step S2-8;

[0129] Step S2-8, the number of iterations t is increased by 1; and when t <t max When , go to step 2-4 to repeat steps 2-4 to 2-8; otherwise, output the optimal position vector parameter of the optimal bat of the tth iteration As the seven photovoltaic cell double diode model parameters of the photovoltaic cell under test.

[0130] The effect of the present invention is verified by comparative experiments below:

[0131] In this comparative test, the IV curve of step 1 contains 11 uniform sampling points, and the size of the battery cell is 25cm 2 , the JV curve also includes 11 sampling points; the value ranges of the dual diode model parameters in step 2-1 are n1∈[1,2], n2∈[2,4], R s ∈[0,1],I ph ∈[10,30],I o1 ∈[10 -9 ,10 -6 ],I o2 ∈[10 -9 ,10 -6 ],R sh ∈[1,1000]; in step 2-2, the number of bats M contained in the bat population is set to 1000, the maximum number of iterations is 1000, and the range of emissivity, loudness, speed, and frequency is [0,1]; in step S2-5, α and γ are taken as 0.9.

[0132] The improved bat algorithm of the present invention is used to estimate the parameters of the dual diode model, and the bat position with the minimum fitness is the optimal parameter of the model. The JV curve fitting results of the improved bat algorithm (IBA) of the present invention and the existing bat algorithm (BA) are shown in Figure 2. Figure 3 As shown in the figure, the BA fitting curve only passes through some sampling points, while the IBA fitting curve passes through all sampling points, achieving good fitting accuracy. In addition, it is calculated that the mean square error of the BA fitting current density is 0.767, while the mean square error of the IBA is 0.004, and the fitting error of the IBA is smaller, thereby confirming the effectiveness of the present invention.

[0133] The present invention is not limited to the above-mentioned specific implementation modes. According to the above-mentioned contents, in accordance with the common technical knowledge and customary means in the field, without departing from the above-mentioned basic technical ideas of the present invention, the present invention can also make other various forms of equivalent modifications, replacements or changes, all of which fall within the protection scope of the present invention.

Claims

1. A photovoltaic cell double diode model identification method based on an improved bat algorithm, which is applicable to the photovoltaic cell under test using a photovoltaic cell double diode model; It is characterized in that include: Step S1, testing and obtaining an IV curve of the photovoltaic cell under standard test conditions, where I represents the port output current of the photovoltaic cell under test, and V represents the port output voltage of the photovoltaic cell under test; Step S2, using an improved bat algorithm to identify seven photovoltaic cell double diode model parameters of the photovoltaic cell under test; The identification process of the improved bat algorithm includes: Step S2-1, setting the parameter value range of the photovoltaic cell under test, including: maximum position vector parameter x max =[n 1max ,n 2max ,R smax ,I phmax ,I o1max ,I o2max ,R shmax ], minimum position vector parameter x min =[n 1min ,n 2min ,R smin ,I phmin ,I o1min ,I o2min ,R shmin ]、Frequency upper limit f max and the frequency lower limit f min ; where n 1max 、n 2max , R smax ,I phmax ,I o1max ,I o2max , R shmax They are the first ideal factor parameter upper limit, the second ideal factor parameter upper limit, the series resistance parameter upper limit, the battery photocurrent parameter upper limit, the first reverse saturation current parameter upper limit, the second reverse saturation current parameter upper limit, and the parallel resistance parameter upper limit; n 1min 、n 2min , R smin ,I phmin ,I o1min ,I o2min , R shmin They are, in order, the first ideal factor parameter lower limit value, the second ideal factor parameter lower limit value, the series resistance parameter lower limit value, the battery photocurrent parameter lower limit value, the first reverse saturation current parameter lower limit value, the second reverse saturation current parameter lower limit value, and the parallel resistance parameter lower limit value; Step S2-2: When the number of iterations t=0, set the maximum number of iterations t max and the number of bats in the bat population M; and the values ​​of bat information of each bat in the bat population are randomly initialized, and the random values ​​obtained by random initialization are recorded as initialization values, wherein the bat information includes: emission rate, loudness, speed, frequency and position vector parameter x = [n1, n2, R s ,I ph ,I o1 ,I o2 ,R sh ], the initialization values ​​are recorded as: r 0 , A 0 、v 0 、f 0 and n1, n2, R s ,I ph ,I o1 ,I o2 , R sh They are the first ideal factor parameter, the second ideal factor parameter, the series resistance parameter, the cell photocurrent parameter, the first reverse saturation current parameter, the second reverse saturation current parameter, and the parallel resistance parameter in the photovoltaic cell double diode model. Step S2-3, for each bat: use the initialization values ​​of the first four parameters in the IV curve and position vector parameter x According to the following formulas 2 to 4, the last three parameters I in the position vector parameter x are o1 ,I o2 , R sh The value of is used to adjust the position: In the formula, X1, X2, The first reverse saturation current parameter I o1 , the second reverse saturation current parameter I o2 , Parallel resistance parameter R sh The value after the position adjustment; A and b both represent matrices. T represents the transposed matrix of matrix A, (A T A) -1 Represents the matrix A T The inverse matrix of A; exp is the exponential power of the natural constant e, N is the number of sampling points of the IV curve, V1,…,V N and I1,…,I N are the port output voltage and port output current from the first sampling point to the Nth sampling point in the IV curve, V t The temperature potential is 25.7 mV under the standard test conditions of the photovoltaic cell; In addition, one of the M bats is randomly selected as the optimal bat of the 0th iteration. The position vector parameter x of the initial optimal bat calculated by formula 2 is recorded as the optimal position vector parameter of the optimal bat of the 0th iteration: Step S2-4, calculate the fitness F(x) of each bat in the bat population according to the following formula 5: In the formula, when the number of iterations t = 0, n1, n2, R s ,I ph ,I o1 ,I o2 , R sh The value of the position vector parameter x is calculated using formula 2 in step S2-3 X1, X2, Calculate; when the number of iterations t≥1, n1, n2, R s ,I ph ,I o1 ,I o2 , R sh The calculation is performed using the value of the position vector parameter x updated after steps S2-5 to S2-7; Step S2-5, adopting the premature convergence control strategy to update the position vector parameter x of each bat in the bat population, that is: first, the M1 bats with the smallest fitness F(x) among the M bats update the position vector parameter x according to the local search method; then, randomly select M2 bats from the remaining M-M1 bats, and update the position vector parameter x through chaotic mutation; finally, the remaining M-M1-M2 bats update the position vector parameter x according to the global search method; wherein, the values ​​of the number of local search bats M1 and the number of chaotic mutation bats M2 are determined according to Formula 6: In the formula, min is the minimum function, ceil is the upward rounding function, and floor is the downward rounding function; In step S2-5, the steps of the local search method are: updating the bat's position vector parameter x according to the following formulas 7 to 9; In the formula, is the firing rate of the i-th bat at the t+1th iteration, is the initialization value of the firing rate of the i-th bat, and γ is the firing rate enhancement coefficient; is the loudness of the i-th bat at the t+1th iteration, is the loudness initialization value of the i-th bat, and α is the loudness attenuation coefficient; is the updated position vector parameter of the i-th bat at the t+1th iteration, is the position vector parameter of the i-th bat calculated by formula 2 in step S2-3, is the random number of the i-th bat at the t-th iteration, t+1 > is the average loudness of all M bats at the t+1th iteration;​ In step S2-5, the chaotic mutation steps are: First, the position vector parameter of the i-th bat is updated at the t-th iteration according to formula 10: Normalize: In the formula, represents the normalized position vector parameter of the i-th bat at the t-th iteration; Then, according to formula 11, the normalized position vector parameters are calculated Piecewise linear chaotic map of : In the formula, represents the mapped position vector parameter of the i-th bat at the t-th iteration; p is the control parameter, and p∈(0,0.5); Finally, according to Formula 12, the mapped position vector parameter Restored to the position vector parameter updated by the i-th bat at the t+1th iteration To update the bat's position vector parameter x: In step S2-5, the steps of the global search method are: updating the bat's position vector parameter x according to formulas 13 to 15; f i =f min +(f max -f min )β i [Formula 13] In the formula, f i represents the sound wave frequency of the i-th bat, f max and f min are the upper and lower frequency limits, respectively, i is the random number of the i-th bat; is the speed of the i-th bat at the t+1th iteration, Initialize the speed of the i-th bat, is the optimal position vector parameter at the tth iteration, is the optimal position vector parameter of the optimal bat selected in step S2-3, is the updated position vector parameter of the i-th bat at the t+1th iteration, is the position vector parameter of the i-th bat calculated by formula 2 in step S2-3; Step S2-6: for each bat whose position vector parameter x is updated in step S2-5, determine whether the parameter in the updated position vector parameter x is out of bounds, that is, determine whether the value θx′ of the parameter exceeds [x min ,x max ] and record the parameters that are judged to be out of bounds as out-of-bounds parameters; The values ​​of out-of-bounds parameters are adjusted as follows: If θx′ is greater than the maximum position vector parameter x max The corresponding parameter θx in max , then the out-of-bounds parameter is adjusted to: If θx′ is less than the minimum position vector parameter x min The corresponding parameter θx in min , then the out-of-bounds parameter is adjusted to: Wherein, θx represents the value after the out-of-bounds parameter is adjusted, and θx′ represents the value before the out-of-bounds parameter is adjusted, that is, the parameter value corresponding to θx in the position vector parameter x updated in step S2-5; θx max and θx min Respectively represent the maximum position vector parameter x max and the minimum position vector parameter x min The parameter corresponding to θx′; k1 and k2 are the first auxiliary judgment parameter and the second auxiliary judgment parameter respectively; mod is the modulus operator; Step S2-7, update the optimal bat in the following manner: First, a random number β1 is generated within a range of values ​​adapted to the emission rate, and a random number β2 is generated within a range of values ​​adapted to the loudness; and the emission rate of the optimal bat at the t+1th iteration is and loudness Update according to Formula 7 and Formula 8; Then, determine whether If the result of the judgment is yes, then: according to Formula 2 to Formula 4, the position vector parameter x of each bat in the tth iteration adjusted by step S2-6 is adjusted, wherein the first four parameters in the position vector parameter x are regarded as the initialization values ​​described in step S2-3 And, according to the position vector parameter x adjusted in this step, the fitness F(x) of each bat is calculated according to the formula 5, and the bat with the smallest fitness F(x) and its position vector parameter x are recorded as the new optimal bat and its position vector parameter The firing rate and loudness of the new optimal bat are respectively and If the judgment result is no, then: directly go to step S2-8; Finally, determine whether and in, represents the fitness of the new optimal bat, represents the fitness of the best bat in the tth iteration; If the result is yes, then: the optimal bat and its position vector parameters of the t+1th iteration are Update to the new optimal bat and its position vector parameters The firing rate of the optimal bat in the t+1th iteration is and loudness Updated to and And, go to step S2-8; If the judgment result is no, then: the optimal bat of the t+1th iteration is maintained as the optimal bat of the tth iteration, and the process goes to step S2-8; Step S2-8, the number of iterations t is increased by 1; and when t <t max , go to step 2-4, otherwise, output the optimal position vector parameter of the optimal bat of the tth iteration As the seven photovoltaic cell double diode model parameters of the photovoltaic cell under test.

Images