CN104778345A - Nonlinear parameter calculation method for simulating aging failure of photovoltaic cell models - Google Patents

Abstract

The invention relates to a nonlinear parameter calculation method for simulating the aging failure of photovoltaic cell models, which includes the following steps: (1) a translucent film blocking method is adopted to decrease irradiance received by a photovoltaic panel in order to simulate the aging failure of the photovoltaic panel, and the relation between the number of translucent film blocking layers and irradiance attenuation is measured; (2) a relation curve between blocking film layer numbers (C) and aging degrees (L) is created; (3) the difference of blocking degrees is utilized to simulate the different aging degrees, a load in a photovoltaic panel circuit is regulated and the voltage and current values of the circuit are acquired, and thereby an I-V output curve of the photovoltaic panel under different irradiances and the different aging degrees is obtained; an external characteristic I-V relation of the photovoltaic panel is established; (4) for the I-V output curve of the photovoltaic panel under the different irradiances and the different aging degrees, optical parameter values are obtained by parameter identification. The nonlinear parameter calculation method can obtain the nonlinear parameter change laws of the aging failure of the photovoltaic cell models.

Description

Nonlinear parameter calculation method for simulating photovoltaic cell model aging fault

Technical Field

The invention belongs to the technical field of photovoltaic cells, and relates to a photovoltaic cell parameter acquisition method.

Background

Due to the pressure of energy crisis and environmental protection, photovoltaic solar energy is valued by all countries due to its cleanness and environmental protection, and has become the trend of future energy development. The failures of photovoltaic cells are classified into aging, splintering, hot spots, and the like. The research on the hot spot problem is developed to the fault location level, and the aging problem of the photovoltaic cell is less relevant research at home and abroad.

Disclosure of Invention

The invention aims to provide a method for solving a change rule of an aging fault parameter of a photovoltaic cell model. The technical scheme of the invention is as follows:

a nonlinear parameter calculation method for simulating photovoltaic cell model aging faults comprises the following steps:

(1) the semi-transparent film shielding mode is adopted to reduce the irradiance received by the photovoltaic cell panel so as to simulate the aging fault of the photovoltaic cell panel, and the more the shielded irradiance is, the deeper the aging degree is; under the same test condition, adopting different numbers of layers of semitransparent films to shield the light receiving surface of the photovoltaic cell panel, and measuring the relation between the number of layers shielded by the semitransparent films and the irradiance attenuation;

(2) the defined degree of ageing (L) is expressed as follows:

degree of aging (L) irradiance attenuation value ÷ irradiance value under standard condition (1000W/m)²)×100％

The aging degree (L) is used for representing the attenuation degree of the light transmittance, the higher the aging degree is, the lower the light transmittance is, and the higher the irradiance attenuation degree is, and a relation curve of the number of shielded thin film layers (C) and the aging degree (L) is established according to the measured relation between the number of shielded layers of the semitransparent thin film and the irradiance attenuation;

(3) simulating different aging degrees by using different shielding degrees, adjusting the load in the photovoltaic cell panel loop, and acquiring loop voltage and current values to obtain I-V output curves of the photovoltaic cell panel under different irradiance and different aging degrees;

(4) obtaining an external characteristic I-V relational expression of the photovoltaic cell panel according to a single diode model of the photovoltaic cell panel and an equivalent circuit thereof:

$I=I_{\mathrm{ph}}-I_0\left\{e\mathrm{xp}\right[\frac{q(V+{\mathrm{IR}}_s)}{\mathrm{AkT}}]-1\}-\frac{V+{\mathrm{IR}}_s}{R_{\mathrm{sh}}}$

wherein,

v-voltage measurement at both ends of photovoltaic cell panel

I-Current in external Circuit of photovoltaic cell Panel

A-diode quality factor

T-Back plate temperature of photovoltaic cell plate

K-Boltzmann constant (1.380X 10-23J/K)

R_s-a battery series resistance

q-electronic charge (1.608X 10-19C)

I_o-reverse saturation current of the diode

I_ph-photovoltaic cell panel photo-generated current

R_sh-parallel resistance of battery

And (3) introducing a Lambert W function to simplify an I-V equation to obtain an explicit expression of the current I of the photovoltaic cell panel:

<math><mrow> <mi>I</mi> <mo>=</mo> <mfrac> <mrow> <msub> <mi>R</mi> <mi>sh</mi> </msub> <mrow> <mo>(</mo> <msub> <mi>I</mi> <mi>ph</mi> </msub> <mo>+</mo> <msub> <mi>I</mi> <mi>o</mi> </msub> <mo>)</mo> </mrow> <mo>-</mo> <mi>V</mi> </mrow> <mrow> <msub> <mi>R</mi> <mi>sh</mi> </msub> <mo>+</mo> <msub> <mi>R</mi> <mi>s</mi> </msub> </mrow> </mfrac> <mo>-</mo> <mfrac> <msub> <mi>AV</mi> <mi>th</mi> </msub> <msub> <mi>R</mi> <mi>s</mi> </msub> </mfrac> <mo>&times;</mo> <mi>W</mi> <mrow> <mo>(</mo> <mfrac> <mrow> <msub> <mi>R</mi> <mi>s</mi> </msub> <msub> <mi>R</mi> <mi>sh</mi> </msub> <msub> <mi>I</mi> <mi>o</mi> </msub> </mrow> <mrow> <msub> <mi>AV</mi> <mi>th</mi> </msub> <mrow> <mo>(</mo> <msub> <mi>R</mi> <mi>sh</mi> </msub> <mo>+</mo> <msub> <mi>R</mi> <mi>s</mi> </msub> <mo>)</mo> </mrow> </mrow> </mfrac> <mi>exp</mi> <mrow> <mo>(</mo> <mfrac> <mrow> <msub> <mi>R</mi> <mi>sh</mi> </msub> <msub> <mi>R</mi> <mi>s</mi> </msub> <mrow> <mo>(</mo> <msub> <mi>I</mi> <mi>ph</mi> </msub> <mo>+</mo> <msub> <mi>I</mi> <mi>o</mi> </msub> <mo>)</mo> </mrow> <mo>+</mo> <msub> <mi>R</mi> <mi>sh</mi> </msub> <mi>V</mi> </mrow> <mrow> <msub> <mi>AV</mi> <mi>th</mi> </msub> <mrow> <mo>(</mo> <msub> <mi>R</mi> <mi>sh</mi> </msub> <mo>+</mo> <msub> <mi>R</mi> <mi>s</mi> </msub> <mo>)</mo> </mrow> </mrow> </mfrac> <mo>)</mo> </mrow> <mo>)</mo> </mrow> <mo>-</mo> <mo>-</mo> <mo>-</mo> <mrow> <mo>(</mo> <mn>2</mn> <mo>)</mo> </mrow> </mrow></math>

wherein, V_thKT/q, W () is LambertW function, and the current I is generated by light_phReverse saturation current I₀Parallel resistor R_shSeries resistance R_sThe I-V curve can be determined by five parameters of the quality factor A of the diode;

(4) for the I-V output curve of the photovoltaic cell panel under different irradiance and different aging degrees, five-dimensional particles x are set_k＝(I_ph,I₀,A,R_s,R_sh) And obtaining the optimal values of the 5 parameters through parameter identification.

As a preferred embodiment, the step (4) adopts a self-adaptive chaotic particle swarm parameter identification method to identify the parameters.

The aging state of the photovoltaic cell is simulated by adopting a method for reducing irradiance of the polyethylene film, the relation between the shielding layer number of the film and the aging degree is quantized, and parameter identification is carried out by using experimental data for simulating aging, so that an expression of a nonlinear parameter in a photovoltaic cell model is obtained, and the nonlinear parameter change rule of the aging fault of the photovoltaic cell model can be obtained.

Drawings

FIG. 1 is a flow chart of the embodiment

FIG. 2 is a view showing the structure of a measuring apparatus

FIG. 3 is a graph showing the relationship between the number of shielding layers and the aging degree of a thin film

FIG. 4 Experimental Main Circuit

FIG. 5 photovoltaic cell model equivalent circuit

FIG. 6 is a flow chart of an actual application of a photovoltaic cell model parameter identification algorithm

FIG. 7 is a graph showing the variation of photo-generated current (Iph) with aging

FIG. 8 is a graph of diode reverse saturation current (Ios) with aging

FIG. 9 is a graph showing the variation of the quality factor (A) of a diode with aging

FIG. 10 is a graph showing the change of the series resistance with aging

Detailed Description

The packaging structure of the photovoltaic cell comprises super white glass, ethylene-vinyl acetate copolymer (EVA), a cell piece, ethylene-vinyl acetate copolymer (EVA) and a back plate. Since the photovoltaic cells are stable in nature and the backsheet mainly plays a supporting role in the photovoltaic module, the effects of ageing of the ultra-white glass and EVA are mainly considered when analyzing the external characterization of the ageing of the photovoltaic cells (the ageing period of the semiconductor material is relatively long).

When the ultra-white glass is exposed to an external environment, the ultra-white glass is aged by external factors such as photo-oxygen, temperature, humidity and the like, and the aging is mainly characterized in that the surface of a glass plate is yellowed, brittle and cracked, the mechanical strength is reduced, and the light transmittance is reduced. The main reason for aging of EVA is that a very small amount of oxygen is left in the finished product of the photovoltaic module, EVA and oxygen are subjected to chemical reaction, and the reaction speed is accelerated along with the enhancement of ultraviolet rays and the increase of temperature. The external characterization of EVA aging is mainly the change of EVA color, and then the change of the light transmittance of the photovoltaic cell is influenced.

The principle of the photovoltaic cell is 'photo-electricity generation', the magnitude of the generated electricity depends on the intensity of sunlight irradiance received by the photovoltaic cell, and when the photovoltaic cell ages, the reflected characteristic is the reduction of the capacity of converting light into electricity, which can also be equal to the reduction of the sunlight irradiance (such as cloudy days), therefore, the first point of the invention is to simulate the aging phenomenon of the photovoltaic cell by reducing the light transmittance of the surface of the photovoltaic cell. Due to the shielding of different layers of the semitransparent film, the light transmittance of the surface of the photovoltaic cell can be reduced to different degrees, and the state of the photovoltaic cell in different aging degrees can be simulated. The invention selects the high-density polyethylene film as a semitransparent film with uniform light transmittance. The high-density polyethylene film has good heat resistance and mechanical strength, small tensile elongation, slightly low light transmittance and strong tear resistance. The slightly lower light transmittance of the high-density polyethylene film meets the design idea of simulating the aging of the photovoltaic cell, and the shielding of the high-density polyethylene film on irradiance is in a stable level due to the small tensile elongation and strong tear resistance.

After the photovoltaic cell is aged, the output electric energy power is reduced. By analyzing the mechanism of the photovoltaic cell, the aging fault of the photovoltaic cell can be simulated by reducing the irradiance received by the photovoltaic cell panel.

Firstly, the implementation process of the method for simulating the aging of the photovoltaic cell by the polyethylene film shielding is described with reference to fig. 1:

the establishment of the relational expression between the number of the shielding films and the aging degree of the photovoltaic cell is the same as the factory standard test conditions (irradiance is 1000w/m ^2, and the temperature is 25 ℃) of the photovoltaic cell. Under standard conditions, the relationship between the number of layers covered by the translucent film material and the irradiance attenuation measured every time the irradiance value is measured every time the number of layers of the shielding film is increased is shown in table 1.

TABLE 1 relationship between number of layers of shielding film and irradiance attenuation

Since the change in light transmittance reflects the change in the degree of aging, we define the degree of aging (L) as follows:

degree of aging (L) irradiance attenuation value ÷ irradiance value under standard condition (1000W/m)²) X 100% degradation (L) is used to denote the degree of attenuation of light transmission, with higher degradation the lower the light transmission (the higher the degree of attenuation of the irradiance). The irradiance attenuation values corresponding to the different film layer numbers in table 1-1 when masked are substituted into the above formula, so as to obtain the relationship curve between the masked film layer number (C) and the aging degree (L), as shown in fig. 3.

Under different aging degrees (different shielding degrees), adjusting the load in the photovoltaic cell loop and collecting the voltage and current values of the loop to obtain I-V output curves of the photovoltaic panel under different conditions.

The method comprises the following specific steps:

the PVC film was first prepared, the level of irradiance reduction under different film masks was measured, and recorded.

And measuring the output voltage and current relation curves of the photovoltaic cells under different simulated aging degrees. Measuring the irradiance of sunlight at 995W/m²Under the condition of 25 ℃, firstly, all 8 layers of films are superposed together and covered on a photovoltaic solar cell panel for shielding, and then experimental equipment is connected as shown in figure 4 or figure 2, and a data acquisition board acquires two voltages measured continuously: the voltage across Rp and r, and the voltage across r. Namely, two pairs of measuring points of the data acquisition card are respectively connected with two ends of Rp and r and two ends of r, and the common point is the cathode of the battery.

Adjusting the resistance value of the slide rheostat to the minimum value of 0 ohm, starting a data measurement system (a data acquisition board is connected with a computer, and a data acquisition system interface of the computer is started to acquire) to start measuring data, increasing the resistance value of the slide rheostat from the minimum value to the maximum value at a proper speed, and then disconnecting the slide rheostat; then quickly removing a layer of shielding film, simultaneously adjusting the resistance value of the slide rheostat to be minimum, then connecting the slide rheostat, and performing the same operation as that in the case of 8-layer filmAdjusting the slide rheostat to the maximum value, and then disconnecting the slide rheostat; and then continuously repeating the steps, reducing the film, adjusting the resistance value of the sliding rheostat to be minimum, connecting the sliding rheostat, adjusting the resistance value of the sliding rheostat to be maximum at a proper speed, and disconnecting the circuit until no film is shielded. This gives the simulated aging test results at this irradiance. Then, the temperature is 25 ℃, and the irradiance is 750W/m respectively²、400W/m²And 240W/m²The same steps are repeated, and the simulated aging experimental data under different irradiances can be obtained after the experiment is completed.

R thus obtained_pAnd the voltage at the two ends of r is the output voltage of the photovoltaic cell, and the resistance value of dividing the voltage at the two ends of r by r is the output current of the photovoltaic cell.

Some obvious error points possibly generated by noise interference and the like in experimental data can be filtered by using a data average filtering algorithm. Because each measured curve is composed of a plurality of points, the calculation amount for parameter calculation of the plurality of data points is overlarge, data extraction is needed, some characteristic points representative of the curve shape are selected to form a characteristic curve, voltage is taken as a basis in the I-V curve, when the voltage is less than 3.5V, all the points in the range of every 0.1V are summed and averaged to obtain an average value point, when the voltage is more than 3.5V, most curves enter a descending area with a larger slope from the beginning, after 3.5V, the points in the range of every 0.05V are summed and averaged, and the average value points are connected into a line, namely, the measured I-V representative curve. And carrying out data filtering and data selection on the remaining 35 groups of voltage and current values according to the method in the same way to obtain an I-V curve.

And then, performing parameter identification by using a self-adaptive chaotic particle swarm parameter identification method. The patent uses a single diode model with the most widely applied photovoltaic solar cell, and five parameters are included, namely, the photoproduction current I_phReverse saturation current I₀Parallel resistor R_shSeries resistance R_sDiode productAnd (4) quality factor A. The physical model is shown in fig. 5.

According to the photovoltaic cell model equivalent circuit, the external characteristic I-V relational expression of the photovoltaic cell can be obtained:

$I=I_{\mathrm{ph}}-I_0\left\{e\mathrm{xp}\right[\frac{q(V+{\mathrm{IR}}_s)}{\mathrm{AkT}}]-1\}-\frac{V+{\mathrm{IR}}_s}{R_{\mathrm{sh}}}---\left(1\right)$

wherein,

v-voltage measurement across photovoltaic modules

I-current in external loop of assembly

A-diode quality factor

T-temperature of back plate of battery

K-Boltzmann constant (1.380X 10-23J/K)

R_s-a battery series resistance

q-electronic charge (1.608X 10-19C)

I_o-diode inversionTo saturation current

I_ph-photovoltaic cell photo-generated current

R_sh-parallel resistance of battery

The current I in the equation can not be processed by an elementary function to obtain an explicit expression, and for this reason, a Lambert W function is introduced to simplify an I-V equation to obtain the explicit expression of the photovoltaic cell current I:

<math><mrow> <mi>I</mi> <mo>=</mo> <mfrac> <mrow> <msub> <mi>R</mi> <mi>sh</mi> </msub> <mrow> <mo>(</mo> <msub> <mi>I</mi> <mi>ph</mi> </msub> <mo>+</mo> <msub> <mi>I</mi> <mi>o</mi> </msub> <mo>)</mo> </mrow> <mo>-</mo> <mi>V</mi> </mrow> <mrow> <msub> <mi>R</mi> <mi>sh</mi> </msub> <mo>+</mo> <msub> <mi>R</mi> <mi>s</mi> </msub> </mrow> </mfrac> <mo>-</mo> <mfrac> <msub> <mi>AV</mi> <mi>th</mi> </msub> <msub> <mi>R</mi> <mi>s</mi> </msub> </mfrac> <mo>&times;</mo> <mi>W</mi> <mrow> <mo>(</mo> <mfrac> <mrow> <msub> <mi>R</mi> <mi>s</mi> </msub> <msub> <mi>R</mi> <mi>sh</mi> </msub> <msub> <mi>I</mi> <mi>o</mi> </msub> </mrow> <mrow> <msub> <mi>AV</mi> <mi>th</mi> </msub> <mrow> <mo>(</mo> <msub> <mi>R</mi> <mi>sh</mi> </msub> <mo>+</mo> <msub> <mi>R</mi> <mi>s</mi> </msub> <mo>)</mo> </mrow> </mrow> </mfrac> <mi>exp</mi> <mrow> <mo>(</mo> <mfrac> <mrow> <msub> <mi>R</mi> <mi>sh</mi> </msub> <msub> <mi>R</mi> <mi>s</mi> </msub> <mrow> <mo>(</mo> <msub> <mi>I</mi> <mi>ph</mi> </msub> <mo>+</mo> <msub> <mi>I</mi> <mi>o</mi> </msub> <mo>)</mo> </mrow> <mo>+</mo> <msub> <mi>R</mi> <mi>sh</mi> </msub> <mi>V</mi> </mrow> <mrow> <msub> <mi>AV</mi> <mi>th</mi> </msub> <mrow> <mo>(</mo> <msub> <mi>R</mi> <mi>sh</mi> </msub> <mo>+</mo> <msub> <mi>R</mi> <mi>s</mi> </msub> <mo>)</mo> </mrow> </mrow> </mfrac> <mo>)</mo> </mrow> <mo>)</mo> </mrow> <mo>-</mo> <mo>-</mo> <mo>-</mo> <mrow> <mo>(</mo> <mn>2</mn> <mo>)</mo> </mrow> </mrow></math>

wherein, V_th＝KT/q， $W\left(\frac{R_s{R}_{\mathrm{sh}}{I}_0}{{\mathrm{AV}}_{\mathrm{th}}(R_{\mathrm{sh}}+R_s)}\exp \left(\frac{R_{\mathrm{sh}}{R}_s(I_{\mathrm{ph}}+I_0)+R_{\mathrm{sh}}V}{{\mathrm{AV}}_{\mathrm{th}}(R_{\mathrm{sh}}+R_s)}\right)\right)$ For the LambertW function, the unknowns are not contained in the brackets because the algorithm is implemented by matlab, where the value can be solved by directly calling the Lambertw () statement.

Thus, the current I is generated by light_phReverse saturation current I₀Parallel resistor R_shSeries resistance R_sFive parameters of the quality factor A of the diode can determine the I-V curve.

Then it is ready toThe method can be used for estimating parameters by applying a group optimization algorithm (self-adaptive chaotic particle swarm algorithm), and setting five-dimensional particles x for any curve_k＝(I_ph,I₀,A,R_s,R_sh) And the global optimal solution obtained after the optimization by using the self-adaptive chaotic particle swarm optimization is the optimal value of the 5 parameters to be identified. The self-adaptive chaotic particle swarm algorithm is realized by a computer matlab, and the method comprises the following specific steps:

step 1: assuming that the number of particle groups is 100, the number of iterations is 150. And setting the value range of each dimension parameter of the particles.

Carrying out chaotic initialization on the particle swarm with the number n by using a Logistic mapping equation (3):

Z_i+1＝μZ_i(1-Z_i),i＝0,1,2,…,μ∈(0,4] (3)

wherein, Z is more than or equal to 0₀≤1，Z_iIs the ith variable. Mu is a control parameter, when mu is 4, the system is in a complete chaotic state, and the chaotic space is [0, 1%]. A random initial value is set for the position and velocity of the particle within the restricted traversal range.

Step 2: taking the position of each particle as its own optimal position p_bestThe optimal particle position in the particle swarm is assigned to the swarm optimal value g_best。

Step 3: and updating the position and the speed of the particle according to the current position and the speed of the particle according to the following updating formulas (4), (5) and (6), and limiting the position and the speed in an allowable value range by utilizing a carrier form.

v_id＝w×v_id+c₁×rand()×(p_id-x_id)+c₂×rand()×(p_gd-x_id) (4)

x_id＝a×x_id+v_id (5)

Wherein, the values of c1 and c2 are usually 2, and the rand () tableIs shown as [0,1 ]]W is a coefficient that holds the original velocity, called the inertial weight, and α is a constraint factor. The traversal ranges of the particle update position variable and the velocity variable are respectively [ X ]_min,X_max]、[V_min,V_max]. In the updating process, if the speed and the position of the particle updating exceed the traversal range, a boundary value is taken.

The weight value in the iterative process is updated by utilizing a strategy of adaptively adjusting the inertia weight w, and the formula is as follows:

<math><mrow> <mi>w</mi> <mo>=</mo> <mrow> <mo>(</mo> <msub> <mi>w</mi> <mi>max</mi> </msub> <mo>-</mo> <msub> <mi>w</mi> <mi>min</mi> </msub> <mo>)</mo> </mrow> <mo>&times;</mo> <mi>exp</mi> <mrow> <mo>(</mo> <mo>-</mo> <msup> <mrow> <mo>(</mo> <mi>&tau;</mi> <mo>&times;</mo> <mfrac> <mi>cur</mi> <mi>Loop</mi> </mfrac> <mo>)</mo> </mrow> <mn>2</mn> </msup> <mo>)</mo> </mrow> <mo>+</mo> <msub> <mi>w</mi> <mi>min</mi> </msub> <mo>-</mo> <mo>-</mo> <mo>-</mo> <mrow> <mo>(</mo> <mn>6</mn> <mo>)</mo> </mrow> </mrow></math>

wherein, w_maxAnd w_minIs the maximum and minimum values of the inertial weight w in equation (4). cur is the current iteration number and Loop is the maximum iteration number. Tau is an empirical coefficient and takes the value of [20, 55%]In the meantime. Because the formula contains a negative exponential part, the cur value is small, the inertia weight w value is large, and the speed and the position of the particle are updated in the whole traversal range at the initial iteration stage of the algorithm; and in the later iteration stage, the cur value is larger, the inertia weight w value is smaller, and the speed and the position of the particle are updated in a small range. Therefore, the adjustment strategy enhances the global search and the layout search of the chaotic particle swarm algorithmCoordination between them.

Step 4: calculating the ordinate (I) of the corresponding point having the same abscissa as each point in the curve_cal) And using formula (7) and the longitudinal coordinate value I of the corresponding point in the original curve_meaThe fitness value of each particle is calculated separately. For each particle, if the current adaptability value of the particle is better than the optimal extreme value of the individual particle, the current adaptability value of the particle is used for replacing the optimal value p of the individual particle_best. Finding out the optimal value of the particle swarm according to the individual optimal value of each particle in the particle swarm, and if the current optimal value of the particle swarm is superior to the historical optimal value, replacing the optimal value g of the particle swarm with the current optimal value_best。

<math><mrow> <mi>f</mi> <mrow> <mo>(</mo> <mi>X</mi> <mo>)</mo> </mrow> <mo>=</mo> <msqrt> <munderover> <mi>&Sigma;</mi> <mrow> <mi>i</mi> <mo>=</mo> <mn>1</mn> </mrow> <mi>N</mi> </munderover> <msup> <mrow> <mo>(</mo> <msub> <mi>I</mi> <mi>cal</mi> </msub> <mo>-</mo> <msub> <mi>I</mi> <mi>mea</mi> </msub> <mo>)</mo> </mrow> <mn>2</mn> </msup> </msqrt> <mo>-</mo> <mo>-</mo> <mo>-</mo> <mrow> <mo>(</mo> <mn>7</mn> <mo>)</mo> </mrow> </mrow></math>

Where X ═ (Iph, Io, a, Rs, Rsh), i.e., the position vector for each particle, represents the parameter values for five cell models. I is_calAnd I_meaThe resulting current value and the actual current measurement value are respectively substituted into equation (7) for the algorithm-identified parameters. Smaller fitness values indicate more accurate identification parameters. And (3) bringing the data of any I-V curve into a group optimization algorithm, setting the cycle times, and continuously optimizing through the algorithm to finally obtain the model parameters which are very close to the true values.

Step 5: judging whether the algorithm reaches the maximum iteration number or meets the convergence condition (the fitness reaches a set value), if so, directly switching to Step7, and if not, executing the next Step.

Step 6: and (4) judging whether the algorithm is early-maturing and converging by using a formula (8). If yes, updating the particle swarm positions according to the formula (5) and the formula (9) to perform chaotic search, and then turning to Step 3. Otherwise, the next step is performed.

<math><mrow> <msup> <mi>&sigma;</mi> <mn>2</mn> </msup> <mo>=</mo> <mfrac> <mn>1</mn> <mi>n</mi> </mfrac> <munderover> <mi>&Sigma;</mi> <mrow> <mi>i</mi> <mo>=</mo> <mn>1</mn> </mrow> <mi>n</mi> </munderover> <msup> <mrow> <mo>(</mo> <mfrac> <mrow> <msub> <mi>f</mi> <mi>i</mi> </msub> <mo>-</mo> <msub> <mi>f</mi> <mi>avg</mi> </msub> </mrow> <mi>f</mi> </mfrac> <mo>)</mo> </mrow> <mn>2</mn> </msup> <mo>,</mo> <mi>f</mi> <mo>=</mo> <mi>max</mi> <mo>[</mo> <mn>1</mn> <mo>,</mo> <mi>max</mi> <mo>|</mo> <msub> <mi>f</mi> <mi>i</mi> </msub> <mo>-</mo> <msub> <mi>f</mi> <mi>avg</mi> </msub> <mo>|</mo> <mo>]</mo> <mo>-</mo> <mo>-</mo> <mo>-</mo> <mrow> <mo>(</mo> <mn>8</mn> <mo>)</mo> </mrow> </mrow></math>

Wherein n is the total number of particles in the particle group. f. of_iIs the fitness value of the ith particle, f_avgIs the current average fitness value of the particle group. Group fitness variance σ²Reflecting the convergence status of the particle swarm if σ²The smaller the size, the more the particle population tends to converge. The invention is sigma²Setting a threshold value when σ²Is less thanIf the threshold value is reached, the algorithm is determined to be early. In order to avoid misjudging the global optimum as premature convergence, an optimum fitness threshold value needs to be set.

When the algorithm is premature, chaotic search is carried out on the position variable of the particles according to the following formula, and the chaotic search is carried out and the updating is carried out:

x_t+1＝X_min+Z_i+1×(X_max-X_min),t＝0,1,2… (9)

wherein x is_i+1As a position variable of the particle, Z_i+1Is a chaotic particle variable, determined by equation (9).

Step 7: and (4) outputting the global optimal solution X to (Iph, Io, A, Rs and Rsh) to obtain an optimal solution of 5 parameters of the model. The algorithm ends.

The flow chart of the algorithm is shown in fig. 3:

and similarly, substituting the data of the remaining 35 curves into an algorithm to obtain parameter identification results under different simulated aging degrees under different irradiances.

The method is characterized in that on the basis of the existing photovoltaic cell universal model (single-diode model), measurement data of the photovoltaic panel aging process are simulated by combining a PVC film material layer-by-layer shielding method, and 5 parameters (photovoltaic cell photo-generated current, diode reverse saturation current, diode quality factor, series resistance and parallel resistance) of the photovoltaic cell model of the photovoltaic cell under different irradiances, temperatures and different aging degrees (different shielding degrees) are solved by applying a group optimization algorithm. The law of the parameters as a function of the degree of ageing, irradiance, is graphically represented below as shown in fig. 4, 5, 6, 7 (at a certain backplate temperature).

Light generation current (I) according to FIG. 4 and Table 1_ph) The relationship between the aging degree and the degree of aging is known as I_phIs approximately linear and inversely related to L. The photo-generated current (I) is defined by considering the influence of aging degree in the formula_ph) Dependence on the relationship among irradiance (S), battery back plate temperature (T), and aging degree (L)The following were used:

I_ph＝a×I_ph,ref×(1-b×L)×S/S_ref

in the formula,

a×I_ph,ref＝c

c is a determined value, the data in the table 1 are substituted, linear fitting is carried out by using a correlation algorithm, and the relation of the photovoltaic cell (under the determined temperature) is obtained:

I_ph＝5.0226×(1-1.122×L)×S/1000

the temperature of the back plate in the experiment is 41 ℃, and the goodness of fit index is as follows:

Goodness of fit:

SSE:1.349

R-square:0.9791

Adjusted R-square:0.9778

RMSE:0.2053

the fitting result of the formula has high goodness of fit with the original data, the adaptability is good, and the original data is well analyzed.

Diode reverse saturation current (I) according to FIG. 5 and Table 1_o) With the change of the aging degree, the reverse saturation current of the diode is approximately in an exponential relationship with the aging degree and is influenced by the irradiance. Defining diode reverse saturation current (I)_o) The relationship between irradiance (S), cell temperature (T), age (L) is as follows:

<math><mrow> <msub> <mi>I</mi> <mi>o</mi> </msub> <mo>=</mo> <mi>a</mi> <mo>&times;</mo> <msub> <mi>I</mi> <mrow> <mi>o</mi> <mo>,</mo> <mi>ref</mi> </mrow> </msub> <mo>&times;</mo> <mrow> <mo>(</mo> <mi>b</mi> <mo>+</mo> <mfrac> <mi>S</mi> <msub> <mi>S</mi> <mi>ref</mi> </msub> </mfrac> <mo>)</mo> </mrow> <mo>&times;</mo> <msup> <mi>e</mi> <mrow> <mi>c</mi> <mo>&times;</mo> <mrow> <mo>(</mo> <mn>1</mn> <mo>-</mo> <mi>L</mi> <mo>)</mo> </mrow> </mrow> </msup> </mrow></math> (T at a specific temperature)

In the formula:

d×I_o,ref＝h

h is a determined value, and then the data in table 1 is substituted into the above formula for data fitting, so as to obtain a relational expression:

<math><mrow> <msub> <mi>I</mi> <mi>o</mi> </msub> <mo>=</mo> <mn>2.775</mn> <mo>&times;</mo> <msup> <mn>10</mn> <mrow> <mo>-</mo> <mn>5</mn> </mrow> </msup> <mo>&times;</mo> <mrow> <mo>(</mo> <mo>-</mo> <mn>0.1342</mn> <mo>+</mo> <mfrac> <mi>S</mi> <mn>1000</mn> </mfrac> <mo>)</mo> </mrow> <mo>&times;</mo> <msup> <mi>e</mi> <mrow> <mn>10.63</mn> <mo>&times;</mo> <mrow> <mo>(</mo> <mn>1</mn> <mo>-</mo> <mi>L</mi> <mo>)</mo> </mrow> </mrow> </msup> </mrow></math>

in the formula, I_oIn mA. The goodness of fit index of the fitting result is as follows:

Goodness of fit:

SSE:0.08914

R-square:0.9423

Adjusted R-square:0.9387

RMSE:0.05278

when the temperature of the photovoltaic cell is a specific value, the variation curve of the diode quality factor (a) along with the irradiance (S) and the aging degree (L) is shown in fig. 6, and it can be roughly seen from the curve that the relationship between the diode quality factor and the aging degree is similar to linear, and has negative correlation and positive correlation with the irradiance.

<math><mrow> <mi>A</mi> <mo>=</mo> <mi>a</mi> <mo>&times;</mo> <msub> <mi>A</mi> <mi>ref</mi> </msub> <mo>&times;</mo> <mrow> <mo>(</mo> <mn>1</mn> <mo>+</mo> <mi>b</mi> <mo>&times;</mo> <mrow> <mo>(</mo> <mn>1</mn> <mo>-</mo> <mi>L</mi> <mo>)</mo> </mrow> <mo>)</mo> </mrow> <mo>&times;</mo> <mrow> <mo>(</mo> <mn>1</mn> <mo>+</mo> <mi>c</mi> <mo>&times;</mo> <mn>1</mn> <mi>n</mi> <mrow> <mo>(</mo> <mfrac> <mi>S</mi> <msub> <mi>S</mi> <mi>ref</mi> </msub> </mfrac> <mo>)</mo> </mrow> <mo>)</mo> </mrow> </mrow></math>

Fitting is performed according to the data in table 1 to obtain the relation:

<math><mrow> <mi>A</mi> <mo>=</mo> <mn>10.07</mn> <mo>&times;</mo> <mrow> <mo>(</mo> <mn>1</mn> <mo>+</mo> <mn>0.8857</mn> <mo>&times;</mo> <mrow> <mo>(</mo> <mn>1</mn> <mo>-</mo> <mi>L</mi> <mo>)</mo> </mrow> <mo>)</mo> </mrow> <mo>&times;</mo> <mrow> <mo>(</mo> <mn>1</mn> <mo>+</mo> <mn>0.1879</mn> <mo>&times;</mo> <mn>1</mn> <mi>n</mi> <mrow> <mo>(</mo> <mfrac> <mi>S</mi> <mn>1000</mn> </mfrac> <mo>)</mo> </mrow> <mo>)</mo> </mrow> </mrow></math>

the goodness of fit index of the fitting result is as follows:

Goodness of fit:

SSE:7.287

R-square:0.9703

Adjusted R-square:0.9664

RMSE:0.4929

FIG. 7 shows the series resistance (R) at a specific photovoltaic cell temperature_s) Along with the variation curves of irradiance and aging degree, the series resistance is approximately in an exponential relationship with the aging degree and is influenced by the irradiance level. Thus, the estimation formula is of the form:

<math><mrow> <msub> <mi>R</mi> <mi>s</mi> </msub> <mo>=</mo> <mi>a</mi> <mo>&times;</mo> <msub> <mi>R</mi> <mrow> <mi>s</mi> <mo>,</mo> <mi>ref</mi> </mrow> </msub> <mo>&times;</mo> <mrow> <mo>(</mo> <mn>1</mn> <mo>+</mo> <mi>b</mi> <mo>&times;</mo> <mfrac> <msub> <mi>S</mi> <mi>ref</mi> </msub> <mi>S</mi> </mfrac> <mo>)</mo> </mrow> <mo>&times;</mo> <mrow> <mo>(</mo> <mn>1</mn> <mo>+</mo> <mi>exp</mi> <mrow> <mo>(</mo> <mi>c</mi> <mo>&times;</mo> <mi>L</mi> <mo>+</mo> <mi>d</mi> <mo>)</mo> </mrow> <mo>)</mo> </mrow> </mrow></math>

fitting the experimental data to obtain the following coefficient relation:

<math><mrow> <msub> <mi>R</mi> <mi>s</mi> </msub> <mo>=</mo> <mn>0.02878</mn> <mo>&times;</mo> <mrow> <mo>(</mo> <mn>1</mn> <mo>+</mo> <mn>2.042</mn> <mo>&times;</mo> <mfrac> <mn>1000</mn> <mi>S</mi> </mfrac> <mo>)</mo> </mrow> <mo>&times;</mo> <mrow> <mo>(</mo> <mn>1</mn> <mo>+</mo> <mi>exp</mi> <mrow> <mo>(</mo> <mn>3.846</mn> <mo>&times;</mo> <mi>L</mi> <mo>-</mo> <mn>1.475</mn> <mo>)</mo> </mrow> <mo>)</mo> </mrow> </mrow></math>

the goodness of fit levels for the fit results were as follows:

Goodness of fit:

SSE:0.2367

R-square:0.9534

Adjusted R-square:0.9489

RMSE:0.08739

although the above fitting formula has a very high degree of matching with experimental data, the result is relatively complex, and in order to simplify the formula and make it more suitable for application in engineering, the relatively simple fitting result is applied as follows:

<math><mrow> <msub> <mi>R</mi> <mi>s</mi> </msub> <mo>=</mo> <mi>a</mi> <mo>&times;</mo> <msub> <mi>R</mi> <mrow> <mi>s</mi> <mo>,</mo> <mi>ref</mi> </mrow> </msub> <mo>&times;</mo> <mrow> <mo>(</mo> <mfrac> <msub> <mi>S</mi> <mi>ref</mi> </msub> <mi>S</mi> </mfrac> <mo>)</mo> </mrow> <mo>&times;</mo> <mrow> <mo>(</mo> <mn>1</mn> <mo>+</mo> <mi>exp</mi> <mrow> <mo>(</mo> <mi>b</mi> <mo>&times;</mo> <mi>L</mi> <mo>+</mo> <mi>d</mi> <mo>)</mo> </mrow> <mo>)</mo> </mrow> </mrow></math>

data are taken in, and fitting results are obtained:

<math><mrow> <msub> <mi>R</mi> <mi>s</mi> </msub> <mo>=</mo> <mn>0.06371</mn> <mo>&times;</mo> <mrow> <mo>(</mo> <mfrac> <msub> <mi>S</mi> <mi>ref</mi> </msub> <mi>S</mi> </mfrac> <mo>)</mo> </mrow> <mo>&times;</mo> <mrow> <mo>(</mo> <mn>1</mn> <mo>+</mo> <mi>exp</mi> <mrow> <mo>(</mo> <mn>3.697</mn> <mo>&times;</mo> <mi>L</mi> <mo>-</mo> <mn>1.25</mn> <mo>)</mo> </mrow> <mo>)</mo> </mrow> </mrow></math>

the goodness of fit is as follows:

Goodness of fit:

SSE:0.3072

R-square:0.9396

Adjusted R-square:0.9358

RMSE:0.09797

experimental results it was also found that if the above formula is continued to be simplified to the following formula:

<math><mrow> <msub> <mi>R</mi> <mi>s</mi> </msub> <mo>=</mo> <mi>a</mi> <mo>&times;</mo> <msub> <mi>R</mi> <mrow> <mi>s</mi> <mo>,</mo> <mi>ref</mi> </mrow> </msub> <mo>&times;</mo> <mrow> <mo>(</mo> <mfrac> <msub> <mi>S</mi> <mi>ref</mi> </msub> <mi>S</mi> </mfrac> <mo>)</mo> </mrow> <mo>&times;</mo> <mi>exp</mi> <mrow> <mo>(</mo> <mi>b</mi> <mo>&times;</mo> <mi>L</mi> <mo>)</mo> </mrow> </mrow></math>

substituting data, the fitting results were as follows:

<math><mrow> <msub> <mi>R</mi> <mi>s</mi> </msub> <mo>=</mo> <mn>0.05531</mn> <mo>&times;</mo> <mrow> <mo>(</mo> <mfrac> <mn>1000</mn> <mi>S</mi> </mfrac> <mo>)</mo> </mrow> <mo>&times;</mo> <mi>exp</mi> <mrow> <mo>(</mo> <mn>2.523</mn> <mo>&times;</mo> <mi>L</mi> <mo>)</mo> </mrow> </mrow></math>

but the goodness of fit is still at the following level:

Goodness of fit:

SSE:0.3684

R-square:0.9275

Adjusted R-square:0.9253

RMSE:0.1057

the fitting result of the formula not only ensures the error precision of the fitting result and the original data, but also considers the practical engineering physical significance and the previous research result, and has very high engineering practical value.

Claims (2)

1. A nonlinear parameter calculation method for simulating photovoltaic cell model aging faults comprises the following steps:

(1) the semi-transparent film shielding mode is adopted to reduce the irradiance received by the photovoltaic cell panel so as to simulate the aging fault of the photovoltaic cell panel, and the more the shielded irradiance is, the deeper the aging degree is; under the same test condition, adopting different numbers of layers of semitransparent films to shield the light receiving surface of the photovoltaic cell panel, and measuring the relation between the number of layers shielded by the semitransparent films and the irradiance attenuation;

(2) the defined degree of ageing (L) is expressed as follows:

degree of aging (L) irradiance attenuation value ÷ irradiance value under standard condition (1000W/m)²)×100％

The aging degree (L) is used for representing the attenuation degree of the light transmittance, the higher the aging degree is, the lower the light transmittance is, and the higher the irradiance attenuation degree is, and a relation curve of the number of shielded thin film layers (C) and the aging degree (L) is established according to the measured relation between the number of shielded layers of the semitransparent thin film and the irradiance attenuation;

(3) simulating different aging degrees by using different shielding degrees, adjusting the load in the photovoltaic cell panel loop, and acquiring loop voltage and current values to obtain I-V output curves of the photovoltaic cell panel under different irradiance and different aging degrees;

(4) obtaining an external characteristic I-V relational expression of the photovoltaic cell panel according to a single diode model of the photovoltaic cell panel and an equivalent circuit thereof:

$I=I_{\mathrm{ph}}-I_0\left\{\exp \right[\frac{q(V+{\mathrm{IR}}_s)}{\mathrm{AkT}}]-1\}-\frac{V+{\mathrm{IR}}_s}{R_{\mathrm{sh}}}$

wherein,

v-voltage measurement at both ends of photovoltaic cell panel

I-Current in external Circuit of photovoltaic cell Panel

A-diode quality factor

T-Back plate temperature of photovoltaic cell plate

K-Boltzmann constant (1.380X 10-23J/K)

R_s-a battery series resistance

q-electronic charge (1.608X 10-19C)

I_o-reverse saturation current of the diode

I_ph-photovoltaic cell panel photo-generated current

R_sh-parallel resistance of battery

And (3) introducing a Lambert W function to simplify an I-V equation to obtain an explicit expression of the current I of the photovoltaic cell panel:

<math> <mrow> <mi>I</mi> <mo>=</mo> <mfrac> <mrow> <msub> <mi>R</mi> <mi>sh</mi> </msub> <mrow> <mo>(</mo> <msub> <mi>I</mi> <mi>ph</mi> </msub> <mo>+</mo> <msub> <mi>I</mi> <mi>o</mi> </msub> <mo>)</mo> </mrow> <mo>-</mo> <mi>V</mi> </mrow> <mrow> <msub> <mi>R</mi> <mi>sh</mi> </msub> <mo>+</mo> <msub> <mi>R</mi> <mi>s</mi> </msub> </mrow> </mfrac> <mo>-</mo> <mfrac> <msub> <mi>AV</mi> <mi>th</mi> </msub> <msub> <mi>R</mi> <mi>s</mi> </msub> </mfrac> <mo>&times;</mo> <mi>W</mi> <mrow> <mo>(</mo> <mfrac> <mrow> <msub> <mi>R</mi> <mi>s</mi> </msub> <msub> <mi>R</mi> <mi>sh</mi> </msub> <msub> <mi>I</mi> <mi>o</mi> </msub> </mrow> <mrow> <msub> <mi>AV</mi> <mi>th</mi> </msub> <mrow> <mo>(</mo> <msub> <mi>R</mi> <mi>sh</mi> </msub> <mo>+</mo> <msub> <mi>R</mi> <mi>s</mi> </msub> <mo>)</mo> </mrow> </mrow> </mfrac> <mi>exp</mi> <mrow> <mo>(</mo> <mfrac> <mrow> <msub> <mi>R</mi> <mi>sh</mi> </msub> <msub> <mi>R</mi> <mi>s</mi> </msub> <mrow> <mo>(</mo> <msub> <mi>I</mi> <mi>ph</mi> </msub> <mo>+</mo> <msub> <mi>I</mi> <mi>o</mi> </msub> <mo>)</mo> </mrow> <mo>+</mo> <msub> <mi>R</mi> <mi>sh</mi> </msub> <mi>V</mi> </mrow> <mrow> <msub> <mi>AV</mi> <mi>th</mi> </msub> <mrow> <mo>(</mo> <msub> <mi>R</mi> <mi>sh</mi> </msub> <mo>+</mo> <msub> <mi>R</mi> <mi>s</mi> </msub> <mo>)</mo> </mrow> </mrow> </mfrac> <mo>)</mo> </mrow> <mo>)</mo> </mrow> <mo>-</mo> <mo>-</mo> <mo>-</mo> <mrow> <mo>(</mo> <mn>2</mn> <mo>)</mo> </mrow> </mrow> </math>

wherein, V_thKT/q, W () is LambertW function, and the current I is generated by light_phReverse saturation current I₀Parallel resistor R_shSeries resistance R_sThe I-V curve can be determined by five parameters of the quality factor A of the diode;

(4) for the I-V output curve of the photovoltaic cell panel under different irradiance and different aging degrees, five-dimensional particles x are set_k＝(I_ph,I₀,A,R_s,R_sh) And obtaining the optimal values of the 5 parameters through parameter identification.

2. The method of claim 1, wherein the step (4) employs an adaptive chaotic particle swarm parameter identification method for parameter identification.