Photovoltaic maximum power point tracking control method based on enhanced salmons swarm optimization algorithm

Abstract

The invention discloses a photovoltaic maximum power point tracking control method based on an enhanced Sabot swarm optimization algorithm, and aims to solve the problems that a plurality of extreme points appear on an output characteristic P-V curve of a photovoltaic array when the photovoltaic array is shielded by buildings, clouds and the like, and a traditional MPPT algorithm is easy to fall into local optimum. The method mainly comprises a rapid positioning stage and an accurate tracking stage, in the rapid positioning stage, the tracking range of the output power of a photovoltaic array in the distributed photovoltaic power generation system is contracted to be near the global MPP of a multi-peak power-voltage characteristic curve output by the photovoltaic array by adopting ISCSO; a P-V characteristic curve output by the photovoltaic array in the contraction range shows a single-peak characteristic; in the accurate tracking stage, variable step size INC is adopted to gradually optimize in a P-V characteristic curve output by the photovoltaic array presenting the unimodal characteristic after rapid positioning so as to obtain higher and more stable photovoltaic output power. By applying the technical scheme, the rapidity and accuracy of tracking the maximum power point by the system can be improved.

Description

Technical Field

[0001] The present invention relates to the technical field of photovoltaic power generation, and in particular to a photovoltaic maximum power point tracking control method based on ESCSO (Enhanced Sand Cat Swarm Optimization). Background Art

[0002] Energy is the material that humans rely on for survival, while the reserves of non-renewable energy such as oil and coal are decreasing sharply, and will cause serious environmental pollution. As a clean and renewable energy, solar energy has a wide range of application prospects. The development and utilization of solar energy has become an important part of my country's energy strategy. In photovoltaic power generation systems, photovoltaic maximum power point tracking (MPPT) technology is a key link, which enables the photovoltaic array to always output the maximum power during operation, thereby improving the energy conversion efficiency of the system.

[0003] At present, distributed photovoltaic power generation systems are widely used in a variety of scenarios. The output characteristics of photovoltaic arrays are mainly affected by ambient temperature and irradiance. They are usually placed in the open air and are easily blocked by high-rise buildings, clouds, etc., which makes the irradiance of the photovoltaic arrays uneven. Traditional MPPT technology has slow tracking speed, large steady-state oscillation, and is prone to misjudgment of disturbance direction when irradiance changes. At the same time, in the case of partial shadow of the photovoltaic array, its output power-voltage characteristic curve will change from a single-peak curve to a multi-peak curve, making the traditional MPPT technology perturbation observation method and conductance increment method easy to fall into local optimality and no longer applicable.

[0004] The meta-heuristic swarm intelligence optimization algorithm has the ability to jump out of the local optimal solution, can solve nonlinear problems, and has strong global search capabilities. Currently, the swarm intelligence optimization algorithm is widely used in the maximum power tracking of solar photovoltaic panels. It can achieve rapid response and global convergence and avoid falling into extreme values.

[0005] The sand cat swarm optimization algorithm is an optimization algorithm based on the swarm intelligence of sand cats. It can quickly find the global optimal solution with a high probability, thereby improving the tracking efficiency of the algorithm. However, it has problems such as slow tracking speed and the algorithm is prone to falling into local optimality in the later stage. It still needs to be fine-tuned in conjunction with traditional MPPT technology. Summary of the invention

[0006] When the photovoltaic array is blocked by buildings, clouds, etc., the output characteristic PV curve of the photovoltaic array will have multiple extreme points, and the traditional MPPT algorithm is prone to fall into the problem of local optimality. The present invention proposes a photovoltaic global maximum power point tracking control method based on ESCSO, which can avoid local optimal values ​​and output the global maximum power point. The method mainly includes: establishing a photovoltaic array model; an MPPT module based on the spiral search-sparrow warning-normal mutation operator-sand cat group optimization algorithm, which is used to shrink the tracking range to the vicinity of the global maximum power point of the multi-peak PV characteristic curve in the rapid positioning stage; an MPPT module based on the variable step size conductance increment method, which is used to continue to gradually optimize the single-peak PV characteristic curve in a small range in the precise tracking stage to obtain a higher and more stable photovoltaic output power. The application of this technical solution can improve the speed and accuracy of the system in tracking the maximum power point.

[0007] In view of the above technical problems, the present invention proposes a photovoltaic maximum power point (MPP) tracking control method based on an enhanced SandCat Swarm Optimization (ESCSO) algorithm, the method comprising a fast positioning stage and a precise tracking stage; wherein:

[0008] 1) In the fast positioning stage, the tracking range of the photovoltaic array output power in the distributed photovoltaic power generation system is narrowed to the vicinity of the global MPP of the multi-peak PV characteristic curve output by the photovoltaic array through the improved Sand Cat Swarm Optimization (ISCSO) algorithm; including:

[0009] 1-1) Initialize the number of sand cat populations and the maximum number of iterations (iter max ) and the position Pos(i) of each sand cat;

[0010] 1-2) Calculate and store the fitness value corresponding to each sand cat position, where the position of the sand cat with the largest fitness value is the current optimal sand cat position (Pos b );

[0011] 1-3) Determine whether the fitness values ​​corresponding to the positions of each sand cat meet the set end conditions of the rapid positioning stage at the same time.

[0012] If the end condition of the fast positioning phase is met, the precise tracking phase begins, and the current optimal sand cat position (Pos b) is the duty cycle D(k-1) for entering the precise tracking stage. At this time, the tracking range of the photovoltaic array output power in the distributed photovoltaic power generation system has been shrunk to the vicinity of the global MPP of the multi-peak PV characteristic curve output by the photovoltaic array. Within the shrunk range, the PV characteristic curve output by the photovoltaic array presents a single-peak PV characteristic curve. If the end condition of the fast positioning stage is not met, the tracking range of the photovoltaic array output power in the distributed photovoltaic power generation system is shrunk to the vicinity of the global MPP of the multi-peak PV characteristic curve output by the photovoltaic array through ISCSO. The specific process is as follows:

[0013] 1-3-1) Initialize nonlinear convergence factor Parameter S M =0.5, a=2.6, b=0.2, where iter c is the current iteration number;

[0014] 1-3-2) Parameter R that divides the attack phase and the search phase in ISCSO,

[0015] R = 2 × r G ×rand(0,1)-r G ,r=r G ×rand(0,1), where r represents the sensitivity range of each sand cat, which is used for position update in the search phase or attack phase in ISCSO; if |R|>0.25, it is the search phase; if |R|≤0.25, it is the attack phase;

[0016] The search phase is as follows: first, randomly select the position of a sand cat in the sand cat population as the current best candidate position (Pos bc ), each sand cat uses the current best candidate position (Pos bc ), the current position of each sand cat (Pos c ) and the current optimal sand cat position (Pos b ) and its sensitivity range (r) to update its position to Pos(t+1);

[0017] Pos(t+1)=e zl ·cos(2πl)·r·(Pos bc (t)-rand(0,1)·Pos c (t))+sin(Pos c (t))+Pos b (t) (1)

[0018] In formula (1),

[0019] The attack phase is as follows: First, each sand cat uses the current optimal sand cat position (Pos b ) and the current position of each sand cat (Pos c ), generate a random position (Pos rnd );

[0020] Pos rnd =|rand(0,1)·Pos b (t)-Pos c (t)| (2)

[0021] Assume that the sensitivity range of the sand cat population is a circle, and use the roulette method to randomly select an angle θ between 0-360° to represent the random movement direction of each sand cat; attack the prey through formula (3), and the sand cat updates its position to Pos(t+1);

[0022] Pos(t+1)=Pos b (t)-r·Pos rnd ·cos(θ) (3)

[0023] After the sand cat updates its position in the search phase or the attack phase, it is determined whether the updated position Pos(t+1) of the sand cat is between 0 and 0.95. If Pos(t+1) is less than 0, Pos(t+1) is 0; if Pos(t+1) is greater than 0.95, Pos(t+1) is 0.95.

[0024] 1-3-3) Calculate the fitness value corresponding to each sand cat's updated position Pos(t+1), and determine whether the fitness value corresponding to each sand cat's position satisfies the set end condition of the fast positioning stage at the same time.

[0025] If the end condition of the fast positioning phase is met, the precise tracking phase begins, and the current optimal sand cat position (Pos b ) is the duty cycle D(k-1) for entering the precise tracking stage. At this time, the tracking range of the photovoltaic array output power in the distributed photovoltaic power generation system has shrunk to the vicinity of the global MPP of the multi-peak PV characteristic curve output by the photovoltaic array. Within the shrunk range, the PV characteristic curve output by the photovoltaic array presents a single-peak PV characteristic curve. If the end condition of the fast positioning stage is not met, the position of the sand cat with the largest fitness value is set to the current optimal sand cat position (Pos b ), the position of the sand cat with the smallest fitness value is set to the current worst sand cat position (Pos w );

[0026] Each sand cat uses the current optimal sand cat position (Pos b) and the current worst sand cat position (Pos w ) and the current position of each sand cat (Pos c ) Update its position Pos(t+1):

[0027]

[0028] Among them, β is a normal distribution random number with mean 0 and variance 1; K is a random number between [-1,1], which is used to control the moving direction and step length of the sand cat; f i is the current fitness value of the sand cat; f g is the global maximum fitness value; f w is the global minimum fitness value; ε is taken as 1e-50 to avoid the denominator being 0;

[0029] Determine whether the updated position Pos(t+1) of the sand cat is between 0 and 0.95. If Pos(t+1) < 0, Pos(t+1) takes 0; if Pos(t+1) > 0.95, Pos(t+1) takes 0.95.

[0030] Randomly select the position of a sand cat in the current sand cat population as the current sand cat position to be updated (Pos cp ), update the current sand cat position to be updated through formula (5):

[0031] Pos cp (t+1)=Pos cp (t)+N·Pos cp (t) (5)

[0032] Where N is a normally distributed random number with a mean of 0 and a variance of 0.25;

[0033] Determine the updated sand cat position Pos cp (t+1) is between 0 and 0.95, if Pos cp (t+1)<0, then Pos cp (t+1) takes 0, if Pos cp (t+1)>0.95, then Pos cp (t+1) takes 0.95;

[0034] 1-3-4) After each sand cat completes the position update, update the number of iterations.

[0035] If the current number of iterations (iter c ) is less than the maximum number of iterations (iter max ), then return to step 1-2) to execute sequentially; otherwise, enter the precise tracking stage, and the current optimal sand cat position (Pos b) is the duty cycle D(k-1) for entering the precise tracking stage. At this time, the tracking range of the photovoltaic array output power in the distributed photovoltaic power generation system has shrunk to the vicinity of the global MPP of the multi-peak PV characteristic curve output by the photovoltaic array. Within the shrunk range, the PV characteristic curve output by the photovoltaic array presents a single-peak PV characteristic curve.

[0036] 2) In the precise tracking stage, the variable step size incremental conductance (INC) method is used to gradually optimize the single peak PV characteristic curve output by the photovoltaic array after rapid positioning, thereby improving the accuracy and speed of power tracking. The process is as follows:

[0037] 2-1) Assume that the photovoltaic output voltage is V and the output current is I, then the photovoltaic output power is:

[0038] P=V*I (6)

[0039] Calculate the change in photovoltaic output voltage dV = V(k)-V(k-1) and the change in photovoltaic output current

[0040] dI=I(k)-I(k-1);

[0041] 2-2) Determine the direction of the disturbance duty cycle (deltaD) based on whether the change in the photovoltaic output voltage (dV) is equal to 0. There are two situations:

[0042] First, when the change in the photovoltaic output voltage (dV) is not equal to 0, the derivative of both sides of formula (6) with respect to V is obtained:

[0043]

[0044] When the system operating point is on the left side of the MPP of the single-peak PV characteristic curve, there is When the system operating point is at the MPP of the single-peak PV characteristic curve, there is When the system operating point is on the right side of the MPP of the single-peak PV characteristic curve,

[0045] Substituting the relationship between the derivative of the photovoltaic output power with respect to the photovoltaic output voltage and 0 into formula (7), we can obtain:

[0046]

[0047] according to The positive or negative value of the current system operating point is used to determine the position of the MPP of the single-peak PV characteristic curve, and the direction of the disturbance duty cycle (deltaD).

[0048] if The sign of the disturbance duty cycle (deltaD) is negative; if The disturbance duty cycle (deltaD) is 0; if Then the sign of the disturbance duty cycle (deltaD) is positive;

[0049] Second, when the change in the photovoltaic output voltage (dV) is equal to 0, if the change in the photovoltaic output current (dI) is greater than 0, the sign of the disturbance duty cycle (deltaD) is negative; if the change in the photovoltaic output current (dI) is equal to 0, the disturbance duty cycle (deltaD) is 0; if the change in the photovoltaic output current (dI) is less than 0, the sign of the disturbance duty cycle (deltaD) is positive;

[0050] 2-3) Use variable step size INC to calculate the disturbance duty cycle (deltaD) through formula (9), and calculate the system duty cycle D(k) through formula (10):

[0051]

[0052] D(k)=D(k-1)+deltaD (10)

[0053] The change of the disturbance duty cycle deltaD continuously disturbs the system duty cycle D(k) to achieve the purpose of accurately tracking the MPP.

[0054] Compared with the prior art, the present invention has the following beneficial effects:

[0055] 1) The present invention aims at the multi-peak characteristics of the output PV characteristic curve of the photovoltaic array in the distributed photovoltaic power generation system under local shading conditions. Considering that the traditional MPPT algorithm may fail to track the MPP, the output power falls into the local optimum and causes unnecessary power loss. First, in the rapid positioning stage, the tracking range is narrowed to the vicinity of the global MPP of the multi-peak PV characteristic curve of the photovoltaic array. Then, in the precise tracking stage, the optimization is gradually sought in a small range of single-peak PV characteristic curves, and finally the maximum output power of the photovoltaic array in the distributed photovoltaic power generation system is obtained quickly and accurately.

[0056] 2) The present invention uses the MPPT controller to achieve the maximum power output of the photovoltaic power generation system, and the algorithm is simple and easy to implement. Compared with other algorithms, it can achieve satisfactory results in terms of convergence speed and convergence accuracy.

[0057] 3) The present invention simulates three typical local shading modes often encountered in distributed photovoltaic power generation systems and adds a stop mechanism to the algorithm. When the boundary shrinks to the global MPP, it can maintain a stable output without generating power oscillations around the global MPP. BRIEF DESCRIPTION OF THE DRAWINGS

[0058] Figure 1-1 It is a block diagram of the photovoltaic MPP tracking control method based on ESCSO proposed in the present invention;

[0059] Figure 1-2 yes Figure 1-1 The flowchart of the photovoltaic MPP tracking control method based on ESCSO is shown;

[0060] Figure 2-1 yes Figure 1-2 Flowchart of the rapid positioning phase shown in;

[0061] Figure 2-2 yes Figure 1-2 Flowchart of the precise tracking phase described in

[0062] Figure 3 It is a schematic diagram of three shielding modes (column shielding, row shielding, and triangle shielding) of the photovoltaic array of the present invention;

[0063] Figure 4 It is a photovoltaic power generation system model based on Boost converter proposed by the present invention;

[0064] Figure 5 This is a comparison chart of the MPP tracking effect between the particle swarm algorithm in occlusion mode and the method of the present invention;

[0065] Figure 6 It is a comparison diagram of the MPP tracking effect of the particle swarm algorithm in the row occlusion mode and the method of the present invention;

[0066] Figure 7 It is a comparison chart of the MPP tracking effects of the triangle occlusion particle swarm algorithm and the method of the present invention. DETAILED DESCRIPTION

[0067] Research materials: The 4×4 series-parallel structure used in the photovoltaic array is as follows Figure 3 As shown, the main circuit is an MPP tracking system based on the Boost circuit. Figure 4 In the simulation model, C in Take 500μF, C out Take 20μF, L take 8.5mH, R load The resistance is 30Ω and the switching frequency of the Boost circuit is 50kHz. The parameters of each component in the simulation model are as follows: short-circuit current I SC =8.55A, open circuit voltage V OC =37.6V, MPP current Im=8.06A, MPP voltage Um=31V. Reference illumination is 1000W / m 2 , the reference temperature is 25℃. The three types of shading of photovoltaic arrays (column shading, row shading, triangle shading) are as follows Figure 3 As shown, the illumination after shielding is 500W / m2 The MPPT controller collects the output voltage and output current of the photovoltaic array, calculates the output power of the photovoltaic array, and uses the ESCSO algorithm to adjust the output impedance of the boost circuit to be equal to the impedance of the photovoltaic array, that is, adjusts the duty cycle of the switch tube to make the impedance match, and transmits the duty cycle to the PWM generator to control the on-off time of the IGBT tube, and finally realizes the tracking of MPP. The ESCSO algorithm in the MPPT controller mainly includes two stages: fast positioning and precise tracking, such as Figure 1-1 As shown, in the rapid positioning stage, ISCSO is used to shrink the tracking range of the photovoltaic array output power in the distributed photovoltaic power generation system to the vicinity of the global MPP of the multi-peak power-voltage characteristic curve output by the photovoltaic array. Within the shrunk range, the PV characteristic curve output by the photovoltaic array presents a single-peak characteristic; in the precise tracking stage, a variable step size INC is used to gradually optimize within the PV characteristic curve output by the photovoltaic array presenting a single-peak characteristic after rapid positioning, so as to obtain a higher and more stable photovoltaic output power. The application of this technical solution can improve the rapidity and accuracy of the system in tracking the maximum power point.

[0068] like Figure 1-2 and Figure 2-1 As shown in the figure, the specific contents of the rapid positioning stage are as follows:

[0069] In this stage, the tracking range of the photovoltaic array output power in the distributed photovoltaic power generation system is mainly reduced to the vicinity of the global MPP of the multi-peak PV characteristic curve output by the photovoltaic array through ISCSO; including:

[0070] 1) Initialize the number of sand cat populations, the corresponding initial positions, and the maximum number of iterations (iter max ), and set the position Pos(i) of the sand cat to a range of 0-0.95;

[0071] 2) Calculate and store the fitness value corresponding to each sand cat, and find the optimal sand cat position (Pos b );

[0072] 3) Determine whether the fitness value of each sand cat meets the end condition of the set fast positioning stage at the same time. If the end condition of the set fast positioning stage is met, it means that it has been tracked to the vicinity of the global optimal value, and you can directly jump out of the loop and output the duty cycle at this time, then enter the precise tracking stage, and the current optimal sand cat position (Pos b) is the duty cycle D(k-1) for entering the precise tracking stage. At this time, the tracking range of the photovoltaic array output power in the distributed photovoltaic power generation system has shrunk to the vicinity of the global maximum power point of the multi-peak PV characteristic curve output by the photovoltaic array. Within the shrunk range, the PV characteristic curve output by the photovoltaic array presents a single-peak PV characteristic curve. If the end condition of the fast positioning stage is not met, the tracking range of the photovoltaic array output power in the distributed photovoltaic power generation system is shrunk to the vicinity of the global maximum power point of the multi-peak PV characteristic curve output by the photovoltaic array through the improved sand cat group optimization algorithm. The specific process is as follows:

[0073] 3-1) Initialization parameter S M =0.5, a new nonlinear convergence factor is introduced in the present invention After verification, a=2.6, b=0.2, iter c is the current iteration number, iter max is the maximum number of iterations set;

[0074] 3-2) Parameter R that divides the attack phase and the search phase in ISCSO,

[0075] Where r represents the sensitivity range of each cat, which is used for position update in the search phase or attack phase in ISCSO; R is the main parameter controlling the difference between attack and search, which is usually set to 1. The present invention has been verified to set the value to 0.25. If |R|>0.25, it is the search phase, otherwise it enters the attack phase.

[0076] When |R|>0.25, the sand cat searches for prey, and each sand cat searches for prey according to the best candidate position (Pos bc ) and the current position (Pos c ) and its sensitivity range (r) to update its own position, and integrate the improved spiral search strategy to enhance the global search ability of the SCSO algorithm in the search phase. The search phase is: randomly select the position of a sand cat in the sand cat population as the current best candidate position (Pos bc ), each sand cat uses the current best candidate position (Pos bc ), the current position of each sand cat (Pos c ) and the current optimal sand cat position (Pos b ) and its sensitivity range (r) to update its position to Pos(t+1);

[0077] Pos(t+1)=e zl ·cos(2πl)·r·(Pos bc (t)-rand(0,1)·Pos c (t))+sin(Posc (t))+Pos b (t) (1)

[0078] In formula (1),

[0079] When |R|≤0.25, the sand cat attacks the prey and enters the attack phase. First, it uses the best position (Pos b ) and the current position (Pos c ) to generate a random position (Pos rnd ).

[0080] Pos rnd =|rand(0,1)·Pos b (t)-Pos c (t)| (2)

[0081] Assuming that the sensitivity range of the sand cat is a circle, the roulette method is used to randomly select an angle θ between 0 and 360° for each sand cat to represent the random movement direction of the sand cat, avoiding the local optimal value; the attack on the prey is realized by formula (3), and the sand cat updates its position to Pos(t+1); the position update formula is as follows:

[0082] Pos(t+1)=Pos b (t)-r·Pos rnd ·cos(θ) (3)

[0083] After the sand cat updates its position during the search phase or the attack phase, it is necessary to determine whether the updated position Pos(t+1) of the sand cat is between 0 and 0.95. If it is out of the range, the position of the sand cat needs to be limited to 0-0.95, that is: if Pos(t+1)<0, then Pos(t+1) takes 0; if Pos(t+1)>0.95, then Pos(t+1) takes 0.95.

[0084] 3-3) Calculate the fitness value corresponding to each updated position Pos(t+1) of each sand cat, and determine whether the fitness value corresponding to each position of the sand cat satisfies the end condition of the set fast positioning stage at the same time.

[0085] If the end condition of the fast positioning phase is met, the precise tracking phase begins, and the current optimal sand cat position (Pos b) is the duty cycle D(k-1) for entering the precise tracking stage. At this time, the tracking range of the photovoltaic array output power in the distributed photovoltaic power generation system has shrunk to the vicinity of the global maximum power point of the multi-peak PV characteristic curve output by the photovoltaic array. Within the shrunk range, the PV characteristic curve output by the photovoltaic array presents a single-peak PV characteristic curve. If the end condition of the fast positioning stage is not met, the position of the sand cat with the largest fitness value is set as the current optimal sand cat position (Pos b ), the position of the sand cat with the smallest fitness value is set to the current worst sand cat position (Pos w );

[0086] Each sand cat uses the current optimal sand cat position (Pos b ) and the current worst sand cat position (Pos w ) and the current position of each sand cat (Pos c ) updates its own position Pos(t+1), integrates the sparrow's alert mechanism, and improves the convergence speed:

[0087]

[0088] Among them, β is a normal distribution random number with mean 0 and variance 1; K is a random number between [-1,1], which controls the moving direction and step length of the sand cat; f i is the current fitness value of the sand cat; f g is the global optimal fitness value; f w is the global worst fitness value; Pos w is the position of the worst sand cat. ε is set to 1e-50 to avoid the denominator being 0.

[0089] Determine whether the updated position Pos(t+1) of the sand cat is between 0-0.95. If Pos(t+1)<0, Pos(t+1) takes 0; if Pos(t+1)>0.95, Pos(t+1) takes 0.95.

[0090] The normal mutation operator is used to improve the algorithm's ability to jump out of the local optimum, and the position of a sand cat in the current sand cat population is randomly selected as the current position of the sand cat to be updated (Pos cp ), update the current sand cat position to be updated through formula (5):

[0091] Pos bc (t+1)=Pos bc (t)+N·Pos bc (t) (5)

[0092] Where N is a normally distributed random number with mean 0 and variance 0.25.

[0093] Determine the updated sand cat position Pos cp (t+1) is between 0 and 0.95, if Pos cp (t+1)<0, then Pos cp (t+1) takes 0, if Pos cp (t+1)>0.95, then Pos cp (t+1) takes 0.95;

[0094] 3-4) After each sand cat completes the position update, one iteration is completed; update the number of iterations, r G , r, R; if the current number of iterations (iter c ) is less than the maximum number of iterations (iter max ), then return to step 2) to execute sequentially and proceed to the next iteration; when the maximum number of iterations (iter max ) or meets the end condition of the fast positioning stage, the iteration stops, enters the precise tracking stage, and switches to the variable step size INC; the current optimal sand cat position (Pos b ) is the duty cycle D(k-1) for entering the precise tracking stage. At this time, the tracking range of the photovoltaic array output power in the distributed photovoltaic power generation system has shrunk to the vicinity of the global maximum power point of the multi-peak PV characteristic curve output by the photovoltaic array. Within the shrunk range, the PV characteristic curve output by the photovoltaic array presents a single-peak PV characteristic curve.

[0095] Entering the precise tracking stage, the variable step size INC is used to gradually optimize the single-peak PV characteristic curve in a small range output by the photovoltaic array after rapid positioning, thereby improving the accuracy and speed of power tracking. Figure 1-2 and Figure 2-2 As shown in Figure 2, the specific contents of the precise tracking stage are as follows:

[0096] First, let the photovoltaic output voltage be V and the output current be I, then the photovoltaic output power is:

[0097] P=V*I (6)

[0098] Calculate the change in photovoltaic output voltage dV = V(k)-V(k-1) and the change in photovoltaic output current

[0099] dI=I(k)-I(k-1);

[0100] Then, the direction of the disturbance duty cycle (deltaD) is determined based on whether the change in the PV output voltage (dV) is equal to 0. There are two situations:

[0101] First, when the change in the photovoltaic output voltage (dV) is not equal to 0, the derivative of both sides of formula (6) with respect to V is obtained:

[0102]

[0103] When the system operating point is on the left side of the MPP of the single-peak PV characteristic curve, there is When the system operating point is at the MPP of the single-peak PV characteristic curve, there is When the system operating point is on the right side of the MPP of the single-peak PV characteristic curve,

[0104] Substituting the relationship between the derivative of the photovoltaic output power with respect to the photovoltaic output voltage and 0 into formula (7), we can obtain:

[0105]

[0106] according to The positive or negative value of the current system operating point is used to determine the position of the MPP of the single-peak PV characteristic curve, and the direction of the disturbance duty cycle (deltaD).

[0107] if The sign of the disturbance duty cycle (deltaD) is negative; if The disturbance duty cycle (deltaD) is 0; if Then the sign of the disturbance duty cycle (deltaD) is positive;

[0108] Second, when the change in the photovoltaic output voltage (dV) is equal to 0, if the change in the photovoltaic output current (dI) is greater than 0, the sign of the disturbance duty cycle (deltaD) is negative; if the change in the photovoltaic output current (dI) is equal to 0, the disturbance duty cycle (deltaD) is 0; if the change in the photovoltaic output current (dI) is less than 0, the sign of the disturbance duty cycle (deltaD) is positive;

[0109] 2-3) Use variable step size INC to calculate the disturbance duty cycle (deltaD) through formula (9), and calculate the system duty cycle D(k) through formula (10):

[0110]

[0111] D(k)=D(k-1)+deltaD (10)

[0112] The change of the disturbance duty cycle deltaD continuously disturbs the system duty cycle D(k) to achieve the purpose of accurately tracking the MPP.

[0113] The simulation example proposed Figure 3The maximum output power of the photovoltaic array under the three shading modes (column shading, row shading, and triangle shading) shown in FIG. 1 is 2972.74 W, 2148.72 W, and 2075.55 W, respectively. The results of tracking the global MPP using the particle swarm optimization (PSO) algorithm and the ESCSO algorithm proposed in the present invention are shown in FIG. Figure 5 , Figure 6 and Figure 7 As shown in the figure, the PSO algorithm tracks the global optimal time of 0.25s, 0.36s, and 0.36s, and the maximum power tracked is 2892.08W, 2147.97W, and 2075.26W, respectively. The ESCSO algorithm proposed in the present invention tracks the global optimal time of 0.20s, 0.17s, and 0.17s, and the maximum power tracked is 2972.74W, 2148.72W, and 2075.55W, respectively. The convergence time is improved by 48.48%, 52.78%, and 52.78%, respectively, and the convergence accuracy is improved by 2.79%, 0.03%, and 0.01%, respectively.

[0114] From the above analysis results of different light intensity examples, it can be seen that the photovoltaic MPP tracking control method based on ESCSO proposed in the present invention has greatly improved the convergence speed, convergence accuracy and global optimization ability compared with the traditional particle swarm algorithm.

[0115] Although the present invention is described above in conjunction with the accompanying drawings, the present invention is not limited to the above-mentioned specific embodiments, which are merely illustrative and non-restrictive. Under the guidance of the present invention, ordinary technicians in the field may make various modifications, improvements, substitutions, etc. without departing from the purpose of the present invention, which are all within the protection of the present invention.

Claims

1. A photovoltaic maximum power point tracking control method based on an enhanced sand cat group optimization algorithm, characterized in that: The method includes a rapid positioning phase and a precise tracking phase; wherein: 1) In the fast positioning stage, the tracking range of the photovoltaic array output power in the distributed photovoltaic power generation system is narrowed to the vicinity of the global maximum power point of the multi-peak PV characteristic curve output by the photovoltaic array through the improved sand cat group optimization algorithm; including: 1-1) Initialize the number of sand cat populations and the maximum number of iterations (iter max ) and the position Pos(i) of each sand cat; 1-2) Calculate and store the fitness value corresponding to each sand cat position, where the position of the sand cat with the largest fitness value is the current optimal sand cat position (Pos b ); 1-3) Determine whether the fitness values ​​corresponding to the positions of each sand cat meet the set end conditions of the rapid positioning stage at the same time. If the end condition of the fast positioning phase is met, the precise tracking phase begins, and the current optimal sand cat position (Pos b ) is the duty cycle D(k-1) for entering the precise tracking stage. At this time, the tracking range of the photovoltaic array output power in the distributed photovoltaic power generation system has shrunk to the vicinity of the global maximum power point of the multi-peak PV characteristic curve output by the photovoltaic array. Within the shrunk range, the PV characteristic curve output by the photovoltaic array presents a single-peak PV characteristic curve. If the end condition of the fast positioning stage is not met, the tracking range of the photovoltaic array output power in the distributed photovoltaic power generation system is shrunk to the vicinity of the global maximum power point of the multi-peak PV characteristic curve output by the photovoltaic array through the improved sand cat group optimization algorithm. The specific process is as follows: 1-3-1) Initialize nonlinear convergence factor Parameter S M =0.5, a=2.6, b=0.2, where iter c is the current iteration number; 1-3-2) Parameter R for dividing the attack phase and the search phase in the improved sand cat swarm optimization algorithm, R = 2×r G ×rand(0,1)-r G ,r=r G ×rand(0,1), where r represents the sensitivity range of each sand cat, which is used for position update in the search phase or attack phase in the improved sand cat swarm optimization algorithm; if |R|>0.25, it is the search phase; if R|≤0.25, it is the attack phase; The search phase is as follows: first, randomly select the position of a sand cat in the sand cat population as the current best candidate position (Pos bc ), each sand cat uses the current best candidate position (Pos bc ), the current position of each sand cat (Pos c ) and the current optimal sand cat position (Pos b ) and its sensitivity range (r) to update its position to Pos(t+1); Pos(t+1)=e zl ·cos(2πl)·r·(Pos bc (t)-rand(0,1)·Pos c (t))+sin(Pos c (t))+Pos b (t) (1) In formula (1), l=Pos bc -Pos b ,k=5; The attack phase is as follows: First, each sand cat uses the current optimal sand cat position (Pos b ) and the current position of each sand cat (Pos c ), generate a random position (Pos rnd ); Pos rnd =|rand(0,1)·Pos b (t)-Pos c (t)| (2) Assume that the sensitivity range of the sand cat population is a circle, and use the roulette method to randomly select an angle θ between 0-360° to represent the random movement direction of each sand cat; attack the prey through formula (3), and the sand cat updates its position to Pos(t+1); Pos(t+1)=Pos b (t)-r·Pos rnd ·cos(θ) (3) After the sand cat updates its position during the search phase or the attack phase, it is determined whether the updated position Pos(t+1) of the sand cat is between 0 and 0.

95. If Pos(t+1) is less than 0, Pos(t+1) is 0. If Pos(t+1) is greater than 0.95, Then Pos(t+1) is 0.95; 1-3-3) Calculate the fitness value corresponding to each sand cat's updated position Pos(t+1), and determine whether the fitness value corresponding to each sand cat's position satisfies the set end condition of the fast positioning stage at the same time. If the end condition of the fast positioning phase is met, the precise tracking phase begins, and the current optimal sand cat position (Pos b ) is the duty cycle D(k-1) for entering the precise tracking stage. At this time, the tracking range of the photovoltaic array output power in the distributed photovoltaic power generation system has shrunk to the vicinity of the global maximum power point of the multi-peak PV characteristic curve output by the photovoltaic array. Within the shrunk range, the PV characteristic curve output by the photovoltaic array presents a single-peak PV characteristic curve. If the end condition of the fast positioning stage is not met, the position of the sand cat with the largest fitness value is set as the current optimal sand cat position (Pos b ), the position of the sand cat with the smallest fitness value is set to the current worst sand cat position (Pos w ); Each sand cat uses the current optimal sand cat position (Pos b ) and the current worst sand cat position (Pos w ) and the current position of each sand cat (Pos c ) Update its position Pos(t+1): Among them, β is a normal distribution random number with mean 0 and variance 1; K is a random number between [-1,1], which is used to control the moving direction and step length of the sand cat; f i is the current fitness value of the sand cat; f g is the global maximum fitness value; f w is the global minimum fitness value; ε is taken as 1e-50 to avoid the denominator being 0; Determine whether the updated position Pos(t+1) of the sand cat is between 0 and 0.

95. If Pos(t+1) < 0, Pos(t+1) takes 0; if Pos(t+1) > 0.95, Pos(t+1) takes 0.

95. Randomly select the position of a sand cat in the current sand cat population as the current sand cat position to be updated (Pos cp ), update the current sand cat position to be updated through formula (5): Pos cp (t+1)=Pos cp (t)+N·Pos cp (t) (5) Where N is a normally distributed random number with a mean of 0 and a variance of 0.25; Determine the updated sand cat position Pos cp (t+1) is between 0 and 0.95, if Pos cp (t+1)<0, then Pos cp (t+1) takes 0, if Pos cp (t+1)>0.95, then Pos cp (t+1) takes 0.95; 1-3-4) After each sand cat completes the position update, update the number of iterations. If the current number of iterations (iter c ) is less than the maximum number of iterations (iter max ), then return to step 1-2) to execute sequentially; otherwise, enter the precise tracking stage, and the current optimal sand cat position (Pos b ) is the duty cycle D(k-1) for entering the precise tracking stage. At this time, the tracking range of the photovoltaic array output power in the distributed photovoltaic power generation system has shrunk to the vicinity of the global maximum power point of the multi-peak PV characteristic curve output by the photovoltaic array. Within the shrunk range, the PV characteristic curve output by the photovoltaic array presents a single-peak PV characteristic curve. 2) In the precise tracking stage, the variable step-size conductance increment method is used to gradually optimize the single-peak PV characteristic curve output by the photovoltaic array after rapid positioning, thereby improving the accuracy and speed of power tracking. The process is as follows: 2-1) Assume that the photovoltaic output voltage is V and the output current is I, then the photovoltaic output power is: P=V*I (6) Calculate the change in photovoltaic output voltage dV = V(k)-V(k-1) and the change in photovoltaic output current dI=I(k)-I(k-1); 2-2) The direction of the disturbance duty cycle (deltaD) is determined based on whether the change in the photovoltaic output voltage (dV) is equal to 0. There are two situations: First, when the change in photovoltaic output voltage (dV) is not equal to 0, Taking the derivative of both sides of formula (6) with respect to V, we get: When the system operating point is on the left side of the maximum power point of the single-peak PV characteristic curve, there is When the system operating point is at the maximum power point of the single-peak PV characteristic curve, there is When the system operating point is on the right side of the maximum power point of the single-peak PV characteristic curve, Substituting the magnitude relationship between the derivative of the photovoltaic output power with respect to the photovoltaic output voltage and 0 into formula (7), we can obtain: according to The positive or negative value of the current system operating point is used to determine the position of the maximum power point of the single-peak PV characteristic curve, and the direction of the disturbance duty cycle (deltaD). if Then the sign of the disturbance duty cycle (deltaD) is negative; if Then the disturbance duty cycle (deltaD) is 0; if Then the sign of the disturbance duty cycle (deltaD) is positive; Second, when the change in photovoltaic output voltage (dV) is equal to 0, If the change in PV output current (dI) is greater than 0, the sign of the disturbance duty cycle (deltaD) is negative; If the change in the PV output current (dI) is equal to 0, the disturbance duty cycle (deltaD) is 0; If the change in PV output current (dI) is less than 0, the sign of the disturbance duty cycle (deltaD) is positive; 2-3) The variable step conductivity increment method is used to calculate the disturbance duty cycle (deltaD) by formula (9), and the system duty cycle D(k) is calculated by formula (10): D(k)=D(k-1)+deltaD (10) The change of the disturbance duty cycle deltaD continuously disturbs the system duty cycle D(k) to achieve the purpose of accurately tracking the maximum power point.

Images