A method for dynamically obtaining output characteristics of a photovoltaic module under carrier motion state

Abstract

The application relates to a photovoltaic module output characteristic dynamic acquisition method under a carrier motion state, which comprises the following steps: 1) collecting acceleration and angular velocity data of the motion carrier in real time; 2) fusing the acceleration and angular velocity data to obtain a motion posture angle model in a carrier coordinate system through a complementary filter; 3) converting the motion posture angle in the carrier coordinate system to a geographic coordinate system; 4) obtaining the light intensity received by the photovoltaic module according to the motion posture angle data, and establishing a photovoltaic module dynamic model according to a photovoltaic module constitutive model; 5) obtaining the output maximum power according to the photovoltaic module dynamic model, and establishing the coupling relationship between the motion parameters of the photovoltaic module and the output maximum power by adopting a bagged random forest weight algorithm; and 6) establishing a photovoltaic module output characteristic dynamic model by adopting a multivariable curve fitting method to obtain the output characteristic of the photovoltaic module under the motion state. Compared with the prior art, the application has the advantages of high precision, strong real-time performance and the like.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of photovoltaic power generation, in particular to a method for dynamically obtaining output characteristics of a photovoltaic module under a carrier motion state. BACKGROUND

[0002] The photovoltaic module is an energy source for photovoltaic power generation. The photovoltaic module is widely used in various fields for power generation, such as offshore floating photovoltaic power station, photovoltaic power station, satellite power supply, photovoltaic electric vehicle and ship, etc. However, the power generation efficiency of the photovoltaic module is affected by the external environment. For example, the offshore floating photovoltaic power station is affected by the flow direction and wind direction, and the change of the posture of the battery panel will change the area of the received light radiation, so the output voltage and current will also change at any time. Therefore, it is necessary to build a dynamic model of the photovoltaic module to obtain the output characteristics of the photovoltaic module under the motion state, so as to maximize the power generation.

[0003] However, most of the current researches on the output characteristics of the photovoltaic module are for the modeling method of the photovoltaic module under the static state, and the influence of the external environment on the output characteristics of the photovoltaic module is not considered, so that the existing modeling method of the photovoltaic module is difficult to be applied to the acquisition of the dynamic characteristics of the photovoltaic module under the motion state. For the photovoltaic module under the motion state, there is a lack of effective and accurate modeling method to unify the motion mechanical parameters and the output electrical parameters of the photovoltaic module. SUMMARY

[0004] The purpose of the present application is to overcome the defects of the prior art and provide a method for dynamically obtaining the output characteristics of a photovoltaic module under a carrier motion state.

[0005] The purpose of the present application can be achieved by the following technical solutions:

[0006] A method for dynamically obtaining the output characteristics of a photovoltaic module under a carrier motion state, comprising the following steps:

[0007] 1) Real-time acquisition of acceleration and angular velocity data of the moving carrier;

[0008] 2) Fusion of the acceleration and angular velocity data by a complementary filter to obtain a motion posture angle model under the carrier coordinate system;

[0009] 3) Conversion of the motion posture angle under the carrier coordinate system to the geographic coordinate system;

[0010] 4) Acquisition of the light intensity received by the photovoltaic module according to the motion posture angle data, and establishment of a dynamic model of the photovoltaic module according to the constitutive model of the photovoltaic module;

[0011] 5) The maximum output power is obtained according to the dynamic model of the photovoltaic module, and a bagged random forest weight algorithm is used to establish the coupling relationship between the motion parameters of the photovoltaic module and the maximum output power;

[0012] 6) A multivariate curve fitting method is used to establish a dynamic model of the output characteristics of the photovoltaic module to obtain the output characteristics of the photovoltaic module under the motion state.

[0013] The inertial measurement unit is installed on the photovoltaic module, and specifically includes an accelerometer and a gyroscope for real-time acquisition of three-axis acceleration and three-axis angular velocity data in the carrier coordinate system.

[0014] In step 2), the complementary filter includes a low-pass filter and a high-pass filter, the measurement signal of the accelerometer passes through the low-pass filter to eliminate high-frequency jitter, and the measurement signal of the gyroscope passes through the high-pass filter to eliminate low-frequency error, and then the motion attitude angle model X(t) at any time is obtained by superposition, that is:

[0015]

[0016] Where Δt is the sampling period, θ(t)、 ψ(t) are the pitch angle, roll angle and heading angle of the carrier motion at t, θ(t-1)、 ψ(t-1) are the pitch angle, roll angle and heading angle of the carrier motion at t-1, ω x (t-1), ω y (t-1), ω z (t-1) are the angular velocity values of the motion carrier measured by the gyroscope at t-1 along the x b , y b , z b axes respectively, θ g , ψ g are the pitch angle, roll angle and heading angle of the accelerometer, H1(s) is the transfer function of the accelerometer measurement signal Y1, H2(s) is the transfer function of the gyroscope measurement signal Y2, is the angle estimation value calculation formula.

[0017] The angle estimation value calculation formula is expressed as:

[0018]

[0019] Where g x (t), g y (t), g z (t) are the angular velocity values of the motion carrier measured by the accelerometer at t along the x b , y b , zb The gravity acceleration value of the shaft.

[0020] In the step 3), the motion attitude angle in the carrier coordinate system is transformed by using the Euler angle method according to the coordinate rotation matrix, and the motion attitude angle data in the geographic coordinate system is obtained, that is,

[0021]

[0022] Wherein, is the angle between the photovoltaic module and the three axes of the geographic coordinate system at time t, C1, C2 and C3 are the coordinate rotation matrix represented by Euler angles, the subscript b represents the carrier coordinate system, and the subscript t represents the geographic coordinate system.

[0023] The step 4) specifically comprises the following steps:

[0024] 41) According to the angle between the light and the three axes of the geographic coordinate system vector And the angle between the photovoltaic module and the three axes of the geographic coordinate system vector The three-axis angle vector γ(t) between the light and the photovoltaic module is obtained, that is,

[0025]

[0026] 42) According to the position relationship between the photovoltaic module and the light, the light intensity G(t) received by the surface of the photovoltaic module is obtained, that is,

[0027] G(t) = |G0(t) γ(t)|

[0028] Wherein, G0(t) is the light intensity of the light at the current moment;

[0029] 43) According to the light intensity G(t) received by the surface of the photovoltaic module, the dynamic model of the photovoltaic module is established through the coupling relationship between the parameters of the photovoltaic module constitutive model and the light intensity and temperature, that is,

[0030]

[0031] Wherein, I pv (t) is the output current of the photovoltaic module at time t, I phr is the short-circuit current of the photovoltaic module under standard conditions, k i is the short-circuit current temperature coefficient, T(t) is the battery temperature at time t, T r is the temperature value under standard conditions, G r is the light intensity value under standard conditions, I or is the diode reverse saturation current under standard conditions, and q is the electronic charge. GEg is the band gap of the photovoltaic module material under standard conditions, A is the diode management factor, K is the Boltzmann constant, U pv (t) is the output voltage of the photovoltaic module, b is a constant, R sr is the equivalent series resistance, R shr is the equivalent parallel resistance.

[0032] The step 5) specifically comprises the following steps:

[0033] 51) Obtain the maximum power point power P m of the photovoltaic module dynamic model, and combine it with the motion parameters M to form the output characteristic parameters T = [P m M] = [P m ω x ω y ω z g x g y g z ] T , and restore the real data after normalization and denormalization of the unified dimension;

[0034] 52) Normalize the output characteristic parameters to obtain the data set under the unified dimension, and obtain the weight value r j between the maximum power point power and the motion parameters by the bagged random forest weight algorithm.

[0035] The step 6) specifically comprises the following steps:

[0036] 61) Draw the maximum power point power P m about each motion state parameter scatter plot, and adopt the least square method to linearly fit the function expression of each motion state parameter and the maximum power point power P m ;

[0037] 62) Based on the multiple linear regression theory, the photovoltaic module output characteristic dynamic model is established, and then:

[0038]

[0039] In the step 61), the function expression of each motion state parameter and the maximum power point power P m is respectively:

[0040]

[0041]

[0042] P m3 = a8·g z 0.03

[0043]

[0044] P m5 =a 13 ·ω y +a 14

[0045]

[0046] wherein a1~a 17 are fitting parameters respectively.

[0047] The photovoltaic module output characteristic dynamic model in the step 62) is specifically:

[0048]

[0049] Compared with the prior art, the present application has the following advantages:

[0050] Firstly, the present application carries out multi-source heterogeneous information sampling and fusion on the motion state parameters of the photovoltaic module and the constitutive model parameters of the photovoltaic module, obtains the dynamic change amount of the light intensity by using coordinate transformation and vector operation, and constructs a photovoltaic module dynamic model based on the constitutive model of the photovoltaic module, thereby providing conditions for realizing the unification of the motion state parameters and the electrical parameters of the photovoltaic module.

[0051] Secondly, the present application obtains the output electrical parameters of the photovoltaic module under the motion state through the obtained photovoltaic module dynamic model, establishes the coupling relationship between the motion state parameters and the output maximum power point power of the photovoltaic module based on the bagged random forest weight algorithm, and constructs a photovoltaic module output characteristic dynamic model by using a multivariate curve fitting method, so as to obtain the output characteristic of the photovoltaic module under the motion state, thereby providing a basis for the maximum power point dynamic tracking control method of the photovoltaic module. BRIEF DESCRIPTION OF DRAWINGS

[0052] Figure 1 It is the method flow principle diagram of the present application.

[0053] Figure 2 It is a complementary filter structure diagram.

[0054] Figure 3 It is a photovoltaic module and light intensity position relationship diagram.

[0055] Figure 4 It is a bagged random forest weight algorithm principle diagram.

[0056] Figure 5 It is a weight diagram of each motion state parameter.

[0057] Figure 6 It is a P m Regarding the acceleration parameter scatter diagram, wherein Fig. (6a) is a P m1As for the acceleration parameter scatter plot, Fig. (6b) is P m2 As for the acceleration parameter scatter plot, Fig. (6c) is P m3 As for the acceleration parameter scatter plot.

[0058] Figure 7 As for the acceleration parameter scatter plot, Fig. (6b) is P m As for the angular velocity parameter scatter plot, Fig. (7a) is P m4 As for the angular velocity parameter scatter plot, Fig. (7b) is P m5 As for the angular velocity parameter scatter plot, Fig. (7c) is P m6 As for the angular velocity parameter scatter plot.

[0059] Figure 8 As for the angular velocity parameter scatter plot.

[0060] Figure 9 As for the angular velocity parameter scatter plot. DETAILED DESCRIPTION

[0061] The present application will be described in detail below with reference to the accompanying drawings and specific embodiments.

[0062] EMBODIMENT

[0063] The present application proposes a photovoltaic module output characteristic dynamic acquisition method under carrier motion state, which realizes the unification of photovoltaic module motion mechanical parameters and output electrical parameters by constructing a multi-source heterogeneous information fusion photovoltaic module output characteristic dynamic model, and obtains the coupling relationship between the maximum output power of the photovoltaic module and the carrier motion state parameters.

[0064] As shown in the figure, the method comprises the following steps: Figure 1

[0065] Step 1: Real-time acquisition of acceleration and angular velocity data of the moving carrier;

[0066] Step 2: Fusion of acceleration and angular velocity data by using a complementary filtering fusion algorithm to obtain the motion attitude angle in the carrier coordinate system;

[0067] Step 3: Conversion of the attitude angle from the carrier coordinate system to the geographic coordinate system;

[0068] Step 4: Obtaining the light intensity received by the photovoltaic module from the attitude angle data, and establishing a photovoltaic module dynamic model from the photovoltaic module constitutive model;

[0069] Step 5: Obtaining the output maximum power from the photovoltaic module dynamic model, and establishing the coupling relationship between the photovoltaic module motion parameters and the output maximum power by using a bagged random forest weight algorithm;

[0070] ​Step 6: Use the multivariate curve fitting method to establish a dynamic model of the output characteristics of the photovoltaic module to obtain the output characteristics of the photovoltaic module under motion.

[0071] The detailed steps are as follows:

[0072] Step 1: Data Acquisition: An inertial measurement unit (IMU) installed on the moving photovoltaic module collects its triaxial acceleration and triaxial angular velocity data in the carrier coordinate system in real time. The IMU includes an accelerometer and a gyroscope. The accelerometer and gyroscope are used to collect the triaxial acceleration of the moving carrier in real time. and triaxial angular velocity data.

[0073] Step 2, Information Fusion: The signals collected by the two sensors are fused using a complementary filtering fusion algorithm to obtain the motion attitude angle of the carrier in the carrier coordinate system.

[0074] The structure of a complementary filter is as follows: Figure 2 As shown, the circuit includes a low-pass filter and a high-pass filter. Let the measurement signals from the accelerometer and gyroscope be Y1 and Y2, respectively, and their corresponding transfer functions be H1(s) and H2(s), respectively. Then we have:

[0075] Y1 = X + U

[0076] Y2 = X + V

[0077]

[0078]

[0079] Where U is the low-frequency noise signal; V is the high-frequency noise signal; and E(s) is the filter gain.

[0080] The precise estimated value obtained by adding them together is:

[0081]

[0082] Formulas for calculating the attitude angle model X(t) and the estimated angle at any given time They are respectively:

[0083]

[0084] Where Δt is the sampling period, in seconds; θ(t), ψ(t) represents the pitch angle, roll angle, and yaw angle of the carrier motion at time t, respectively, and θ(t-1), ψ(t-1) represents the pitch angle, roll angle, and yaw angle of the carrier motion at time t-1, respectively, and ω x (t-1), ω y (t-1), ωz (t-1) are angular velocity values of the motion carrier measured by the gyroscope respectively in x b , y b , z b axes at t-1 moment, θ g , ψ g are estimated values of the pitch angle, roll angle and heading angle respectively calculated by the accelerometer, H1(s) is a transfer function of the accelerometer measurement signal Y1, H2(s) is a transfer function of the gyroscope measurement signal Y2, is an angle estimation value calculation formula, then:

[0085]

[0086] wherein g x (t), g y (t), g z (t) are gravity acceleration values of the motion carrier measured by the accelerometer respectively in x b , y b , z b axes at t moment.

[0087] Step 3, coordinate conversion: obtain coordinate rotation matrices C1, C2, C3 by using the Euler angle method:

[0088] The motion attitude angle of the carrier in the carrier coordinate system is transformed into the motion attitude angle in the geographic coordinate system by using the coordinate conversion method, then:

[0089]

[0090] wherein, is the angle between the photovoltaic module and the three axes of the geographic coordinate system at t moment, the subscript b represents the carrier coordinate system, and the subscript t represents the geographic coordinate system.

[0091] Step 4, dynamic modeling of the photovoltaic module, specifically including the following steps:

[0092] (1) as shown in Figure 3 , the three-axis angle vector γ(t) between the light intensity and the photovoltaic module is obtained by using vector operation relationship, then:

[0093]

[0094] wherein, is the three-axis angle vector between the light intensity and the geographic coordinate system, is the three-axis angle vector between the photovoltaic module and the geographic coordinate system.

[0095] (2) as shown in Figure 3 , the light intensity G(t) received by the surface of the photovoltaic module is obtained, then:

[0096]

[0097] (3) Through the coupling relationship between the parameters of the photovoltaic module constitutive model and the light intensity and temperature, a dynamic model of the photovoltaic module is established.

[0098] The constitutive model of the photovoltaic module is:

[0099]

[0100] The dynamic model of the photovoltaic module is:

[0101]

[0102] Wherein, I pv is the output current of the photovoltaic module, A; I ph is the photo-generated current of the photovoltaic module, A; I o is the reverse saturation current of the diode, A; U pv is the output voltage of the photovoltaic module, V; R s is the equivalent series resistance, Ω; R sh is the equivalent parallel resistance, Ω; q is the electronic charge, 1.6×10 -19 C; K is the Boltzmann constant, 1.38×10 -23 J / K; A is the diode ideality factor; T is the absolute temperature, K; I phr is the short-circuit current of the photovoltaic module under standard conditions, A; k i is the temperature coefficient of the short-circuit current; T (t) is the temperature of the battery at t, ℃; T r is the temperature value under standard conditions, 25℃; G (t) is the light intensity at t, W / m 2 ; G r is the light intensity value under standard conditions, 1000W / m 2 ; I or is the reverse saturation current of the diode under standard conditions, A; E g is the energy band width of the photovoltaic module material under standard conditions, 1.12 eV; b is about equal to 0.218.

[0103] Step 5, weight calculation, specifically comprising the following steps:

[0104] (1) The maximum power point power is obtained from the dynamic model of the photovoltaic module, and is combined with the motion parameter group to form the output characteristic parameter T=[P m M]=[P m ω x ω y ω z g x gy g z ] T , and the real data is restored by using normalization and inverse normalization to unify the dimension; wherein the normalization and inverse normalization formulae are as follows:

[0105]

[0106] t i = t' i (t max -t min )+t min

[0107] wherein, t max , t min are the maximum value and the minimum value in the arbitrary column data; t i is the original input value; t' i is the normalized value of t i ; t i is the inverse normalized value of t' i .

[0108] (2) The normalized output characteristic parameters obtain the data set in the unified dimension, and the weight value expression r = [r1 r2... r n ] T , as shown in Fig. 2, the weight values are obtained by the bagging random forest weight algorithm. The bagging random forest weight algorithm calculation formula, the weight values of the maximum power point power and each motion parameter are as follows: Figure 4

[0109]

[0110]

[0111] wherein, n oob (k) is the sample number of the kth out-of-bag data set; P m,t (k) , and are the real value, the predicted value before disturbance and the predicted value after disturbance of the tth sample in the kth out-of-bag data set, in this example, the weight values are specifically as follows:

[0112] r = [12.8 3.8 73.6 5.3 0.7 3.8] T

[0113] Step 6, curve fitting, specifically comprising the following steps:

[0114] ​(1) as shown in Table 1, by drawing P m Regarding each motion state parameter scatter diagram, the function expression of single motion state parameter and P m is obtained by using the least square linear fitting.

[0115] Table 1 P m The function table of each motion state parameter fitting

[0116]

[0117] The specific function expression of each single motion state parameter and P m is obtained, respectively:

[0118]

[0119]

[0120] P m3 = 62.843g z 0.03

[0121]

[0122] P m5 = -0.0048ω y + 59.486

[0123]

[0124] (2) Based on the multiple linear regression theory, a dynamic model of photovoltaic module output characteristics is established

[0125] The dynamic model of photovoltaic module output characteristics is finally obtained as:

[0126]

[0127] The present application is aimed at the photovoltaic module under the carrier motion state for experimental simulation, such as under the given train running speed and track vibration parameter, the dynamic output characteristics of photovoltaic module are obtained by the modeling method, and the experimental results are shown in Figures 5-9 .

[0128] Finally, it is necessary to point out that: the above is only the preferred example of the present application, and does not limit the present application, any modification, equivalent replacement and improvement made in the spirit and principle of the present application should be included in the protection scope of the present application.

Claims

1. A method for dynamically obtaining output characteristics of a photovoltaic module under carrier motion state, characterized in that, The method comprises the following steps: 1) collecting acceleration and angular velocity data of the moving carrier in real time; 2) fusing the acceleration and angular velocity data to obtain a motion attitude angle model in a carrier coordinate system by a complementary filter; 3) converting the motion attitude angle in the carrier coordinate system to a geographic coordinate system; 4) obtaining the light intensity received by the photovoltaic module according to the motion attitude angle data, and establishing a photovoltaic module dynamic model according to a photovoltaic module constitutive model; 5) obtaining the output maximum power according to the photovoltaic module dynamic model, and establishing a coupling relationship between the motion parameters of the photovoltaic module and the output maximum power by using a bagged random forest weight algorithm; 6) establishing a photovoltaic module output characteristic dynamic model by using a multivariate curve fitting method to obtain the output characteristics of the photovoltaic module under the motion state. The complementary filter in the step 2) comprises a low-pass filter and a high-pass filter, the measurement signal of the accelerometer passes through the low-pass filter to eliminate high-frequency jitter, the measurement signal of the gyroscope passes through the high-pass filter to eliminate low-frequency error, and then the signals are superposed to obtain a motion attitude angle model at any time Then, wherein, is the sampling period, , , are respectively t the pitch angle, roll angle, heading angle of the vehicle motion at time t, , , are respectively t- the pitch angle, roll angle, heading angle of the vehicle motion at time t, , , are respectively t- the angular velocity values of the vehicle motion measured by the gyroscope at time t along the , , axes, , , are respectively the pitch angle, roll angle, heading angle estimation values of the accelerometer, is the transfer function of the accelerometer measurement signal , is the transfer function of the gyroscope measurement signal , is the angle estimation value calculation formula; The step 5) specifically comprises the following steps: 51) obtaining the maximum power point power from the dynamic model of the photovoltaic assembly and combining with the motion parameters to form the output characteristic parameters and restoring the real data after unified dimension with normalization and denormalization; 52) Normalization of output characteristic parameter to obtain data set in unified dimension, and obtain weight value between maximum power point power and motion parameter by bagging random forest weight algorithm .

2. The method of claim 1, wherein the output characteristics of the photovoltaic module are dynamically obtained under the carrier motion state. The photovoltaic module is provided with an inertial measurement unit, specifically comprising an accelerometer and a gyroscope for collecting three-axis acceleration and three-axis angular velocity data in the carrier coordinate system in real time.

3. The method of claim 1, wherein the output characteristics of the photovoltaic module are dynamically obtained in a carrier motion state. The angle estimation value solving formula The expression is: wherein , , is t the gravity acceleration values of the motion carrier measured by the accelerometer respectively in , , axes.

4. The method of claim 1, wherein the output characteristics of the photovoltaic module are dynamically obtained in a carrier motion state. In the step 3), the motion attitude angle in the carrier coordinate system is transformed to obtain the motion attitude angle data in the geographic coordinate system by using the Euler angle method according to the coordinate rotation matrix, that is, wherein, is t the angle between the photovoltaic module and the three axes of the geographic coordinate system at the moment, , , are respectively the coordinate rotation matrices expressed in Euler angles, the subscript b denotes the carrier coordinate system and the subscript t denotes the geographic coordinate system.

5. The method of claim 1, wherein the output characteristics of the photovoltaic module are dynamically obtained in a carrier motion state. The step 4) specifically comprises the following steps: 41) from the vector of the angles between the light ray and the three axes of the geographic coordinate system and the vector of the angles between the photovoltaic module and the three axes of the geographic coordinate system the vector of the angles between the light ray and the photovoltaic module then: 42) According to the relationship between the photovoltaic module and the light position, the light intensity received by the surface of the photovoltaic module is obtained Then, there is: wherein, is the light intensity of the light ray at the current time instant; 43) according to the intensity of the light received by the surface of the photovoltaic module By coupling the parameters of the photovoltaic module constitutive model with the intensity of the light and the temperature, a dynamic model of the photovoltaic module is established, then: wherein, is t the output current of the photovoltaic module at time t, is the short circuit current of the photovoltaic module at standard conditions, k i is the short circuit current temperature coefficient, is t the cell temperature at time t, is the temperature value at standard conditions, is the light intensity value at standard conditions, is the diode reverse saturation current at standard conditions, q is the electron charge, is the photovoltaic module material band gap at standard conditions, A is the diode ideality factor, K is the Boltzmann constant, is the photovoltaic module output voltage, b is a constant, is the equivalent series resistance, is the equivalent parallel resistance.

6. The method of claim 1, wherein the output characteristics of the photovoltaic module are dynamically obtained under the carrier motion state. The step 6) specifically comprises the following steps: 61) by plotting maximum power point power For each scatter plot of the motion state parameters, a linear function expression of each motion state parameter and maximum power point power Pmp is obtained by least square method. For each scatter plot of the motion state parameters, a linear function expression of each motion state parameter and maximum power point power Pmp is obtained by least square method. 62) based on the multiple linear regression theory, a photovoltaic module output characteristic dynamic model is established, that is, 。 7. The method of claim 6, wherein the output characteristics of the photovoltaic module are dynamically obtained under the carrier motion state. In step 61), each motion state parameter is related to the maximum power point power. The function expressions are as follows: wherein are fitting parameters, respectively.

8. The method of claim 6, wherein the output characteristics of the photovoltaic module are dynamically obtained in a carrier motion state. In the step 62), the photovoltaic module output characteristic dynamic model is specifically as follows: 。

Images