Multi-Extremum Photovoltaic Power Generation Maximum Power Point Tracking Method Based on Honey Badger Algorithm

Abstract

The present invention relates to a multi-extremum photovoltaic maximum power point tracking method based on the honey badger algorithm, belonging to the field of photovoltaic power generation control. It includes calculating the current power in real time by using the output voltage and current of the overall photovoltaic battery string, mining and tracking to find the global maximum power point through the honey badger optimization algorithm, and at the same time giving the judgment basis for environmental changes. When the external environment changes, this method can also adaptively track to the new global maximum power point, and there is no power fluctuation after stabilization. By using the method described in the present invention, only the output voltage and current of the overall photovoltaic string need to be collected to find the global maximum power point, without any other information, not relying on a model, and applicable to photovoltaic strings with different parameters under the condition of power and voltage level matching, with strong versatility, and having good dynamic response and steady-state performance for both static environment conditions and dynamic environment conditions.

Description

Technical Field

[0001] The present invention belongs to the technical field of photovoltaic power generation control, and particularly relates to a multi-extremum photovoltaic maximum power point tracking method based on the honey badger algorithm. Background Technique

[0002] With the progress of technology and industrial upgrading, the demand for energy in today's society is increasing day by day. Under the background of carbon neutrality and energy conservation and emission reduction, the demand for green environmental protection restricts the further development of traditional power generation methods. Solar energy is one of the most abundant renewable energy sources on the earth today. Photovoltaic power generation has developed rapidly with its characteristics of green environmental protection, low maintenance cost, no need for fuel, no moving parts, easy installation and distributed deployment, and abundant reserves, becoming a research hotspot in various fields and being widely applied in various countries.

[0003] The output of a photovoltaic cell has strong non-linear characteristics. Its P-V and I-V curves both have a maximum power point and are affected by light intensity and temperature. Under different environmental conditions, the position of its maximum power point is different and changes in real time with the environment. Under uniform light and the same temperature conditions, the output characteristic curve of the photovoltaic cell shows a single peak; under non-uniform light and temperature conditions, its output characteristic curve shows multiple peaks.

[0004] In order to achieve the maximum power output of the photovoltaic battery string, generally, a DC / DC converter is connected to the load or the DC bus behind it, and the maximum power point tracking (MPPT) is realized by adjusting the duty ratio of the DC / DC converter to change the equivalent resistance of the external circuit of the photovoltaic battery string.

[0005] Traditional algorithms such as the perturbation observation method and the conductance increment method cannot track the global maximum power point under partial shading conditions. Most model methods require prior knowledge of the number of series-connected photovoltaic cells and rely on the accuracy of the model used. The neural network method needs to train different networks for different specifications of photovoltaic cells, and its practicability and versatility are slightly poor. [Ge Qiang, Sun Tao, etc. Photovoltaic MPPT control system based on sliding mode variable structure-global comparison composite algorithm [P]. Jiangsu Province: CN116991195A, 2023-11-03.], A photovoltaic MPPT control system based on a sliding mode variable structure-global comparison composite algorithm is proposed in this patent. This method is complex to implement and requires additional temperature and light sensors to achieve MPPT, which will increase additional costs in engineering applications, and its experimental effect also highly depends on the accuracy of the model used. [Xu Hengshan, Zhao Mingyang, etc. A photovoltaic MPPT control method based on improved mayfly algorithm [P]. Hubei Province: CN116880650A, 2023-10-13.], This patent uses an improved mayfly algorithm to achieve photovoltaic MPPT control. This method still requires a large number of perturbation steps under partial shading conditions. The short tracking time is partly due to the very short MPPT period. The computing power of the processor in engineering applications is limited, and the actual hardware cannot support too low an MPPT period, further reducing the practicability of reducing the perturbation steps. [Zhang Wenfeng, Wu Tong, etc. An MPPT method based on improved grey wolf optimization algorithm [P]. Sichuan Province: CN116845953A, 2023-10-03.], This patent presents an MPPT method based on an improved grey wolf optimization algorithm, which also has a large number of perturbation steps and greater power fluctuations, making it not conducive to practical applications. Summary of the Invention

[0006] Aiming at the defects existing in the existing MPPT technology, the purpose of the present invention is to track the global maximum power point as fast as possible and with as high accuracy as possible under partial shading conditions, and be able to adaptively adjust the working point according to environmental changes, so as to improve the utilization efficiency of photovoltaic energy.

[0007] To achieve the above object, the technical solution adopted by the present invention is: A multi-extremum photovoltaic power generation maximum power point tracking method based on the Honey Badger Algorithm (HBA), including the following steps:

[0008] (1) Initialize the number of particles N, the duty cycle of each individual [d0, d1, d2,..., d i ,..., d N-1 ,..., d i ,..., I N-1 ;

[0009] (2) Sample the voltage and current values output by the current photovoltaic cell and calculate the current power;

[0010] (3) Determine whether the global maximum power point has been found;

[0011] (4) If the global maximum power point is not found in step (3), update the honey attraction degrees of each individual [I0, I1, I2, …, I i , …, I N-1 , …, d i , …, d N-1 ;

[0012] (5) If the global maximum power point has been found in step (3), stabilize at the duty cycle d gbest corresponding to the global maximum power point and calculate the fluctuating power at all times to determine whether a new maximum power point is generated due to a change in the external environment;

[0013] (6) If a new maximum power point is generated in step (5), re-initialize the duty cycles d i of each individual and return to step (3) to re-perform the maximum power point tracking.

[0014] Furthermore, the duty cycle d i of the i-th individual in step (1) is uniformly initialized according to the number of particles N and the upper and lower limits of the duty cycle d max , d min as follows:

[0015]

[0016] where 0 ≤ i < N and i is an integer.

[0017] Furthermore, in step (3), it is determined whether the maximum power point has been found according to whether Δd sum is less than the given value ε, and the calculation formula is as follows:

[0018]

[0019] where Δd sum represents the degree of proximity between the duty cycles of each individual, N represents the number of particles, d i represents the duty cycle of the i-th individual, and d i+1 represents the duty cycle of the (i + 1)-th individual.

[0020] Even further, the honey attraction degree I i of the i-th individual and the duty cycle d i of the i-th individual are updated according to the duty cycle d i+1 of the next adjacent individual and the duty cycle d prey corresponding to the current group maximum power point, and the calculation formula is as follows:

[0021]

[0022] When r1 ≤ 0.5, it is in the excavation mode:

[0023] d i-new = d prey + F × β × I i × d prey + F × r3 × α × (d i - d prey )

[0024] When r1 > 0.5, it is in the honey mode:

[0025] d i-new = d prey + F × r4 × α × (d i - d prey )

[0026] In the formula, γ is a constant, and its default value is 0.2. d i-new represents the updated value of d i . r1, r2, r3, and r4 are all random numbers between 0 and 1. F randomly takes 1 or -1. α and β are both constants greater than or equal to 1, and their default values are 6 and 2 respectively.

[0027] Furthermore, in step (5), when the current power P new deviates from the power P gbest at the previously recorded maximum power point by a percentage ΔP of the absolute value greater than the upper limit |ΔP| max , the time t max when ΔP is greater than the upper limit |ΔP| gx is started to be calculated. If within an MPPT cycle, ΔP is greater than the deviation upper limit |ΔP| max for x% of the time, it is considered that the environment has changed and the position of the maximum power point has changed. The calculation formula is as follows:

[0028]

[0029] t gx > x% × T MPPT

[0030] In the formula, P new is the current power, P gbest is the power at the previously recorded maximum power point, ΔP is the percentage by which P new deviates from P gbest , t gx is the time when ΔP is greater than the upper limit |ΔP| max , T MPPT is the MPPT cycle, and x ∈ (0, 100).

[0031] The present invention is directed to a photovoltaic power generation system composed of a photovoltaic battery string, a DC / DC converter, and a load. By introducing the honey badger optimization algorithm to solve the maximum power point tracking problem under the condition of partial shading of photovoltaic panels, the overall output voltage and current of the photovoltaic battery string are used to calculate the current power in real time, and the global maximum power point is found by excavation and tracking. At the same time, the determination basis for environmental changes is given. When the external environment changes, this method can also adaptively track the new global maximum power point, and there is no power fluctuation after stabilization. By using the method described in the present invention, only the overall output voltage and current of the photovoltaic string need to be collected to find the global maximum power point, without any other information, independent of the model, and applicable to photovoltaic strings with different parameters under the condition of power and voltage level matching, with strong versatility, and having good dynamic response and steady-state performance for both static and dynamic environmental conditions. Brief Description of the Drawings

[0032] Figure 1 Schematic diagram of the structure of a multi-photovoltaic panel series photovoltaic power generation system;

[0033] Figure 2 Output power-voltage curve of the photovoltaic string under local shading environmental condition 1;

[0034] Figure 3 Output power-voltage curve of the photovoltaic string under local shading environmental condition 2;

[0035] Figure 4 Flowchart of the maximum power point tracking method for multi-extremum photovoltaic power generation based on the honey badger algorithm;

[0036] Figure 5 Multi-extremum photovoltaic maximum power point tracking process based on HBA under environmental condition 1;

[0037] Figure 6 Multi-extremum photovoltaic maximum power point tracking process based on HBA under environmental condition 2; Detailed Description of the Invention

[0038] In order to more clearly illustrate the technical solution of the present invention, the present invention will be described in detail below with reference to the drawings and embodiments.

[0039] The structure of the multi-photovoltaic panel series photovoltaic power generation system is as Figure 1 shown. The number of series-connected photovoltaic panels is 3, numbered ①②③ from top to bottom. In the experiment, three satellite sailboard power array simulators of model DC176302 are used instead, and two local shading conditions, environmental condition 1 and environmental condition 2, are set, and their parameters are shown in Table 1 and Table 2 respectively. For the three series-connected photovoltaic panels, the overall output power-voltage curve under environmental condition 1 is as Figure 2As shown, the global maximum power point is 169.25W; the power-voltage curve of the overall output under environmental condition 2 is as Figure 3 shown, and the global maximum power point is 104W. The DC / DC converter selects a synchronous rectification structure to reduce the converter loss, and the two switching tubes conduct complementarily. The MPPT controller selects TMS320F28335. The DC bus voltage is set to 150V.

[0040] Table 1 Photovoltaic cell parameters under environmental condition 1

[0041]

[0042] Table 2 Photovoltaic cell parameters under environmental condition 2

[0043]

[0044] The flow chart of the maximum power point tracking method for multi-extremum photovoltaic power generation based on the honey badger algorithm is as Figure 4 shown.

[0045] In step (1), initialize the number of particles N = 4, and set the upper and lower limits of the duty cycle d max , d min to 0.1 and 0.75 respectively, initialize the duty cycle of each individual [d0, d1, d2, d3] = [0.2625, 0.425, 0.5875, 0.75], initialize the honey attractiveness of each individual [I0, I1, I2, I3] = [0.2625, 0.425, 0.5875, 0.75]; initialize the algorithm flag bit mode, set 1 to indicate that the maximum power point tracking is in progress, and set 0 to indicate that the global maximum power point has been found; the particle number is u; the interrupt count variable is counter; the interrupt frequency is f INT ; the MPPT frequency f MPPT = 5.

[0046] In step (2), collect the voltage and current values of the overall output of the current photovoltaic cell string, calculate the current power, and record the power at the current moment and the previous moment.

[0047] In step (3), judge whether the maximum power point has been found according to whether Δd sum is less than the given value ε (here ε = 1.0×10 -4 ), and its calculation formula is as follows:

[0048]

[0049] In step (4), if the algorithm flag bit mode is 1 and the maximum power point tracking is in progress, and within the interrupt time when the duty cycle perturbation should be performed, that is

[0050]

[0051] Then, the individual duty cycles d0, d1, d2, and d3 are output to the duty cycle d for the actual operation of the converter in sequence. current , based on whether the current power is greater than the previously recorded maximum power. If the current power is greater, then update the duty cycle corresponding to the maximum power point of the current population (i.e., the optimal position d of the population prey ) to the current duty cycle; otherwise, d prey remains unchanged.

[0052] Then, update the honey attractiveness I of each individual i , and update the duty cycle d of each individual i . The calculation formula is as follows:

[0053]

[0054] When r1 ≤ 0.5, it is in the excavation mode:

[0055] d i-new = d prey + F × β × I i × d prey + F × r3 × α × (d i - d prey )

[0056] When r1 > 0.5, it is in the honey mode:

[0057] d i-new = d prey + F × r4 × α × (d i - d prey )

[0058] In the formula, γ is a constant, taking 0.2, and d i-new represents the value after d i is updated. r1, r2, r3, and r4 are all random numbers between 0 and 1. F randomly takes 1 or -1. α and β are both constants greater than or equal to 1, and their default values are 6 and 2 respectively.

[0059] According to step (5), if the algorithm flag bit mode is 0 and the global maximum power point has been found, then the current duty cycle is the duty cycle d corresponding to the global maximum power point gbest . The duty cycle of the DC / DC converter stabilizes at d gbest and operates without change, and no longer updates the honey attractiveness I of each individual i and the duty cycle d of each individual i . However, it is necessary to record the percentage of power deviation ΔP at all times. If ΔP is greater than the deviation upper limit for x% of the time within an MPPT cycle

[0060] |ΔP|max , it is considered that the environment has changed and the position of the maximum power point has changed. Here, x is taken as 90, and the upper limit of deviation |ΔP| max is taken as 10%, and the MPPT period T MPPT is 0.2 s. The discrimination formula for environmental change is as follows:

[0061]

[0062] t ΔP>10％ > 90% × 0.2 s

[0063] In the formula, P new is the current power, and P gbest is the power at the maximum power point recorded previously.

[0064] From step (6), if a new maximum power point is generated in step (5), then the duty cycle d of each individual i is re-initialized, and step (3) is returned to re-perform maximum power point tracking. Otherwise, it continues to operate at the duty cycle d corresponding to the global maximum power point gbest .

[0065] The multi-extremum photovoltaic maximum power point tracking process based on HBA under environmental condition 1 is as Figure 5 shown. The global maximum power point is 169.25 W. The global maximum power point power found by the method of the present invention is 166.8 W, the voltage is 83 V, the current is 2.01 A, and the tracking efficiency is 98.6%; the tracking time is 2 s, one MPPT cycle is 0.2 s, and a total of 10 MPPT cycles are used. The power-voltage curve of the overall output under environmental condition 2 is as Figure 6 shown. The global maximum power point power is 104 W. The global maximum power point found by the method of the present invention is 102 W, the voltage is 39 V, the current is 2.62 A, and the tracking efficiency is 98.1%; the tracking time is 2.4 s, one MPPT cycle is 0.2 s, and a total of 12 MPPT cycles are used.

[0066] The experimental results prove that the method used in the present invention can quickly track the global maximum power point under the environmental conditions of local shading of the photovoltaic string, and has good tracking accuracy and tracking speed.

[0067] The present invention can be implemented in other specific forms without departing from its spirit or essential characteristics. The described embodiments are considered to be illustrative rather than restrictive in all aspects. For example:

[0068] 1) The structure of the selected DC-DC converter;

[0069] 2) The values of the constants in the duty cycle and honey attractiveness update formulas;

[0070] 3) The selection of various parameters of the photovoltaic string in the experiment, the position of the photovoltaic operating point when the environment changes, etc.

[0071] Accordingly, the scope of the present invention is indicated by the appended claims rather than the above description. All changes that fall within the meaning and scope of equivalent technical solutions of the claims are included in its scope.

Claims

1. A multi-extremum photovoltaic power generation maximum power point tracking method based on the honey badger algorithm, characterized in that: (1) Initialize the number of particles N and the duty cycle of each particle [d0, d1, d2, …, d i ,…,d N-1 ], each individual honey attraction [I0,I1,I2,…,I i ,…,I N-1 ], i is the particle number, the duty cycle of the i-th individual is d i According to the number of particles N and the upper and lower limits of the duty cycle d max d min Uniform initialization, the calculation formula is as follows: where 0 ≤ i < N and i is an integer; (2) Sample the voltage and current values output by the current photovoltaic cell and calculate the current power; (3) According to Δd sum Whether it is less than a given value ε determines whether the maximum power point has been found. The calculation formula is as follows: In the formula, Δd sum Indicates the closeness between the duty cycles of each individual, N represents the number of particles, d i represents the duty cycle of the ith individual, d i+1 represents the duty cycle of the i+1th individual; (4) If the global maximum power point is not found in step (3), update the attractiveness of each individual honey [I0,I1,I2,…,I i ,…,I N-1 ], update the duty cycle of each individual [d0, d1, d2, …, d i ,…,d N-1 ], the honey attraction of the i-th individual is I i , the duty cycle d of the i-th individual i , according to the duty cycle d of the next adjacent body i+1 And the duty cycle d corresponding to the current group maximum power point prey Update, the calculation formula is as follows: When r1 ≤ 0.5, it is in the excavation mode: d i-new =d prey +F×β×I i ×d prey +F×r3×α×(d i -d prey ) When r1 > 0.5, it is in the honey mode: d i-new =d prey +F×r4×α×(d i -d prey ) In the formula, γ is a constant, and its default value is 0.2, d i-new Indicates d i After the update, r1, r2, r3, and r4 are all random numbers between 0 and 1, F is randomly 1 or -1, and α and β are constants ≥ 1, with default values ​​of 6 and 2 respectively; (5) If the global maximum power point has been found in step (3), then stabilize at the duty cycle d corresponding to the global maximum power point gbest Run and calculate the fluctuating power at all times to determine whether the external environment has changed and a new maximum power point has been generated; (6) If step (5) generates a new maximum power point, then the duty cycle d i Reinitialize and return to step (3) to perform maximum power point tracking again.

2. A multi-extreme value photovoltaic power generation maximum power point tracking method based on honey badger algorithm as claimed in claim 1, characterized in that: In step (5), when the current power P new Deviation from the previously recorded maximum power point power P gbest The percentage of absolute value ΔP is greater than the upper limit |ΔP| max When ΔP is greater than the upper limit |ΔP|, the calculation starts max Time t gx , if x% of the time in an MPPT cycle ΔP is greater than the upper limit of deviation |ΔP| max , it is considered that the environment has changed and the position of the maximum power point has changed. The calculation formula is as follows: t gx >x%×T MPPT Where P new is the current power, P gbest is the maximum power point power recorded previously, ΔP is P new Deviation from P gbest The percentage of gx When ΔP is greater than the upper limit |ΔP| max Time, T MPPT is the MPPT period, x∈(0,100).

Images