Light storage system cooperative intelligent control method considering local shadow condition

Abstract

The invention relates to the technical field of photovoltaic power generation system control, in particular to a light storage system cooperative intelligent control method considering a local shadow working condition, which combines the output power of a photovoltaic system with the charging and discharging states of an energy storage battery under the condition that a photovoltaic array is in local shading. And distributing the generated power of the photovoltaic system to the energy storage system or the load. When a photovoltaic system is in a local shading condition, a manta ray foraging optimization algorithm is combined with an improved self-adaptive variable-step-size incremental conductance method, a global optimal point is found through an intelligent algorithm, and then stable tracking is realized through the improved self-adaptive variable-step-size incremental conductance method. A self-adaptive droop control method based on a state of charge (SOC) and a state of health (SOH) is adopted. According to the method, an intelligent algorithm is combined, the maximum power tracking problem when the photovoltaic system is in partial shading is solved, the system stability is effectively maintained, meanwhile, collaborative management of the photovoltaic system and the energy storage system is achieved, and the service life of an energy storage battery is effectively prolonged.

Description

Technical Field

[0001] The present invention relates to the field of photovoltaic power generation system control technology, and in particular to a photovoltaic storage system collaborative intelligent control method taking into account local shading conditions. Background Art

[0002] Nowadays, the use of renewable energy for power generation has become the most important trend in the global power generation field. Renewable energy has become an important source of electricity for mankind and is becoming increasingly important. The proportion of renewable energy power generation represented by wind power and photovoltaic power generation using distributed power generation is gradually increasing.

[0003] Under partial shading conditions, the output power of the photovoltaic array exhibits multi-peak fluctuations. When using the traditional maximum power tracking method, it is easy to fall into the local optimum and cannot track the true global optimum, thus affecting reliable power supply. Therefore, it is necessary to combine intelligent algorithms on this basis to reduce the adverse effects on the power system.

[0004] The structure of a DC microgrid is simpler than that of an AC microgrid, which can reduce device costs and power losses, thereby improving the energy utilization efficiency of the system. The DC microgrid control layer only needs to control the output of each unit based on the bus voltage, making it easier to control distributed power sources and loads, and having a lower failure rate.

[0005] However, compared with AC microgrid, it lacks inertia link, so its anti-interference ability is weaker, and it is necessary to maintain power balance at all times to maintain bus voltage stability.

[0006] Photovoltaic power generation units need to be combined with energy storage units to achieve a balance between supply and demand of load electricity consumption. Research on the coordinated control between photovoltaic and energy storage DC microgrids, microgrids and AC power grids to achieve stable operation of DC microgrids is of great significance to improving the renewable energy absorption capacity and optimizing the energy structure. Summary of the invention

[0007] The purpose of the present invention is to solve the shortcomings existing in the prior art and to propose a collaborative intelligent control method for a photovoltaic storage system taking into account local shading conditions. Through the autonomous control of the converter, the working modes of the photovoltaic system and the energy storage system are regulated in real time. When multiple batteries are connected in parallel, the working mode of the energy storage battery can be dynamically adjusted based on the changes in the battery's state of charge SOC and state of health SOH.

[0008] In order to achieve the above object, the present invention adopts the following technical solutions:

[0009] A method for cooperative intelligent control of a photovoltaic storage system taking into account local shading conditions comprises the following steps:

[0010] When the photovoltaic array is in partial shadow, its output characteristic curve will have multiple peak points. The traditional maximum power point tracking technology is prone to fall into the local optimum, resulting in tracking failure and a significant reduction in the output efficiency of the photovoltaic system.

[0011] Therefore, the traditional variable step-size conductance increment method is improved to an adaptive variable step-size conductance increment method, and the "Manta Ray Foraging Optimization Algorithm MRFO" is introduced to combine it.

[0012] The Manta Ray Foraging Optimization (MRFO) algorithm simulates the foraging process of manta rays in the ocean and mathematically describes the way in which individual manta rays update their positions, thereby realizing the search for the optimal solution in a complex solution space. Among them, the Manta Ray Foraging Optimization algorithm has three foraging behaviors, including chain foraging, spiral foraging, and tumbling foraging:

[0013] Step 1.1) During the chain predation process, the moving direction and step length of the next position of the manta ray individual are determined by the current optimal solution and the previous individual position. The mathematical model of this position update method is as follows:

[0014]

[0015] In the formula, represents the position of the ith individual in the t+1th generation in the d dimension; Respectively represent the positions of the t-th generation, the i-th and the i-1-th individuals in the d dimension; r is a random number uniformly distributed on [0,1]; a is a factor related to r; represents the position of the best individual of the tth generation in the dth dimension; N represents the number of individuals, Indicates the distance between the t-th generation, the i-th individual and the optimal position in the dimension in d dimension;

[0016] When the MRFO algorithm is used to track the maximum power point of the photovoltaic system, the output power of the photovoltaic array is used as the objective function (fitness), the duty cycle corresponding to the photovoltaic array voltage is used as the position of the individual, and the duty cycle of the voltage corresponding to the global maximum power point of the photovoltaic array is used as the optimal solution of the population;

[0017] Step 1.2) When a manta ray finds a prey, due to the existence of the chain predation method, the manta ray is also affected by the previous individual in the process of moving to the current spiral. The mathematical model of this position update method is as follows:

[0018]

[0019] Where T is the total number of iterations; r 1 is a random number uniformly distributed on [0,1]; β is 1 The relevant factor; r is the random number used in step 1.1);

[0020] When t / T≤rand, rand is a random number uniformly distributed in the interval [0,1]. The mathematical equation describing the spiral motion of manta rays can be defined as:

[0021]

[0022] In the formula, represents the random position of the tth generation and the dth dimension; Ub d Indicates the upper bound of the variable value; Lb d represents the lower bound of the variable value; r is the random number used in step 1.1), Ub d -Lb d Indicates the size of the variable interval;

[0023] Step 1.3) In the rolling predation, the manta ray uses the current optimal solution as the rolling fulcrum and rolls to the other side that is a mirror image of its current position. The mathematical model is expressed as follows:

[0024]

[0025] In the formula, r 2 、r 3 They are all random numbers uniformly distributed in the interval [0,1];

[0026] Step 1.4) On this basis, the standard deviation of all individuals and the distance between individuals and the optimal individual are introduced into the inertia weight of each individual to judge the convergence speed of the individual. The formula is as follows:

[0027]

[0028] In the formula, ω 0 is a constant, indicating the fixed fluctuation degree of algorithm convergence; ω i Indicates the convergence weight of the i-th individual in the current dimension; λ 1 and λ 2 are all constants; σ represents the standard deviation of all contemporary individual positions in the current dimension; Indicates the distance of the ith individual in the tth generation from the current optimal individual position in the dth dimension;

[0029] Each individual is constantly iterating and updating its speed and position, gradually moving closer to the optimal position. Therefore, when many individuals reach a certain extreme point, it can be considered that the extreme point is the global maximum power point. At this time, σ will be a constant close to 0.

[0030] Therefore, when the individual reaches the vicinity of the global maximum point, it is considered that |ω i -ω 0| is a minimum value close to 0, at which point |ω i -ω 0 |≤λ 1 a 1 +λ 2 a 2 ; In the formula, a 1 and a 2 They are all constants close to 0;

[0031] When the external environment changes, the global maximum power point of the photovoltaic system will change, so the algorithm needs to re-track the maximum power point and set the algorithm restart condition to Where P 1 represents the power after the environmental change, and P represents the power before the environmental change;

[0032] Step 1.5) The process of manta ray foraging optimization algorithm is as follows:

[0033] Step 1.51. Set relevant parameters and initialize the population;

[0034] Step 1.52. Calculate the initial fitness value;

[0035] Step 1.53. Determine whether the condition rand < 0.5 is met. If so, perform spiral foraging. If not, perform chain foraging.

[0036] Step 1.54. Calculate the fitness value and update the optimal position;

[0037] Step 1.55. Perform tumbling foraging and update the position;

[0038] Step 1.56. Calculate the fitness value and update the optimal position;

[0039] Step 1.57. Determine whether the end condition is met. If so, output the duty cycle corresponding to the maximum power point. Otherwise, repeat steps 1.52 to 1.57 to continue iterative updating.

[0040] Preferably, the traditional conductivity increment method determines whether to add or subtract the duty cycle according to the size of dP / dU. When the photovoltaic cell works at the maximum power point, dP / dU=0; when it is on the left side of the maximum power point, dP / dU>0; when it is on the right side of the maximum power point, dP / dU<0;

[0041] Step 2.1) Adaptive variable step size algorithm should ensure the step size ΔD on this basis 1 It is a monotonically increasing function of dP / dU. When the operating voltage of the photovoltaic array is close to the maximum power point voltage, ΔD 1 Small, can effectively reduce the fluctuation near the maximum power point, improve the accuracy of the algorithm, and its derivative ΔD 1' is larger, so it can improve the ΔD 1 When the operating voltage of the photovoltaic array is far away from the maximum power point voltage, ΔD 1 is larger, thus speeding up the dynamic response of the algorithm, and its derivative ΔD 1 ' is small, which can ensure that the algorithm can be adjusted quickly in a short time. Here, the adaptive change formula of step size is proposed:

[0042] dU=U K -U K-1

[0043] dI=I K -I K-1

[0044] P=U K I K

[0045]

[0046] In the formula, I k-1 Represents the current value at the previous moment; I k Represents the current value at the current moment; U k-1 Represents the voltage value at the previous moment; U k represents the voltage value at the current moment; dU represents the change in voltage; dI represents the change in current; P is the power at the current moment; a is about Factor of ΔD 1 Represents each time The step size for resizing; k 1 and c 1 is a constant used to and ΔD 1 Scale it to a suitable size;

[0047] Step 2.2) In the conductance increment method, when dU = 0, it is necessary to further determine the size of dI. When dI is small, it is considered that this time is close to the maximum power point, ΔD 2 Small, can effectively reduce the fluctuation near the maximum power point, improve the accuracy of the algorithm, and its derivative ΔD 2 ' is larger, so it can improve the ΔD 2 When dI is large, it is considered that it is far away from the maximum power point, ΔD 2 is larger, thus speeding up the dynamic response of the algorithm, and its derivative ΔD 2 'Small, which can ensure that the algorithm can be adjusted quickly in a short time;

[0048] Because the size and change speed of dI are different from dP / dU, the above formula is not applicable. To address this shortcoming, the above formula is modified and ΔD is proposed.2 The formula for dI is:

[0049] b=k 2 ·dI

[0050]

[0051] Where b is a factor about dI; ΔD 2 represents the step size adjusted according to the size of dI; k 2 and c 2 are all constants;

[0052] In step 2.3), the optimal position obtained by the manta ray foraging optimization algorithm is used as the initial duty cycle of the improved adaptive variable step size conductance increment method, and then the maximum power tracking is performed; if the algorithm restart condition mentioned in step 1.4) is met, it means that the external lighting conditions have changed, that is, the maximum power point of the photovoltaic array has changed, so the manta ray foraging optimization algorithm is returned to search for the best result.

[0053] Preferably, since the photovoltaic system DC / DC converter will cause certain disturbances to the DC bus voltage, an energy storage battery is used to smooth the DC bus voltage disturbance, and the following formula is added to the energy storage battery droop control function:

[0054]

[0055] Among them, U in0 Represents the input voltage of the photovoltaic cell DC / DC converter at the last moment, U in1 Represents the input voltage of the photovoltaic cell DC / DC converter at the current moment, D 0 Represents the duty cycle of the photovoltaic cell DC / DC converter at the last moment, D 1 Represents the duty cycle of the photovoltaic cell DC / DC converter at the current moment, Represents the no-load DC reference voltage, ΔU out Represents the disturbance caused by the photovoltaic cell DC / DC converter to the DC bus voltage, U N Represents the nominal voltage of the DC bus;

[0056] Step 3.1) The droop control function expression is: In the formula, is the reference value of the output voltage of the droop control of the i-th energy storage module, P i is the output power of the i-th energy storage module, R i is the droop coefficient of the i-th energy storage module;

[0057] Taking the dual energy storage in parallel as an example, the droop control function expression is: Since the voltage drop caused by droop control is low, it can be considered that Combined with the droop control expression, the output power ratio between energy storages can be obtained: Where P 1 Represents the output power of the first energy storage battery, P 2 Represents the output power of the second energy storage battery, R 1 Represents the droop factor of the first energy storage battery, R 2 represents the droop factor of the second energy storage battery;

[0058] Step 3.2) In a system with multiple energy storages in parallel, the improved droop coefficient R of the converter of the i-th energy storage system i The calculation formula is as follows:

[0059]

[0060] Among them, R i is the droop coefficient of the i-th energy storage system converter; R 0 is the initial droop coefficient; SOC i represents the i-th energy storage battery; Represents the average state of charge of k energy storage batteries; ΔSOC i represents the difference between the state of charge of the i-th energy storage battery and the average state of charge; Represents the average health status of k energy storage batteries; ΔSOH i Represents the difference between the health status of the i-th energy storage battery and the average health status; β 1 , α 1 , β 2 , α 2 are all constants; n is the energy storage module equilibrium speed adjustment factor;

[0061] Step 3.3) Order BATT i Indicates the state of the i-th energy storage battery;

[0062]

[0063] When the energy storage battery works in discharge mode, the ratio of the output power between the energy storage batteries is:

[0064]

[0065] When the energy storage battery works in charging mode, the ratio of the output power between the energy storage batteries is:

[0066]

[0067] Therefore, the output power of the two energy storage batteries can be expressed as:

[0068]

[0069] Among them, P sum Represents the total output power of the two energy storage batteries;

[0070] When the energy storage module is in the discharge mode, the energy storage battery with a larger SOC and SOH has a smaller adaptive droop coefficient, the converter discharge current is larger, and the output power is larger; the energy storage battery with a smaller SOC and SOH has a larger adaptive droop coefficient, the converter discharge current is smaller, and the output power is smaller, so that the battery status reaches the same level. When the energy storage module is in the charging mode, the energy storage battery with a larger SOC and SOH has a larger adaptive droop coefficient, the converter discharge current is smaller, and the output power is smaller; the energy storage battery with a smaller SOC and SOH has a smaller adaptive droop coefficient, and its converter discharge current is larger, and the output power is larger. The above method can achieve SOC balancing of the energy storage module and adjust the discharge depth of the energy storage module at the same time, thereby achieving SOH balancing.

[0071] Compared with the prior art, the present invention has the following beneficial effects:

[0072] 1. The present invention combines the manta ray foraging optimization algorithm with the improved adaptive variable step-size conductance increment method to achieve maximum power tracking for partially shaded photovoltaic arrays, thereby improving energy utilization efficiency.

[0073] 2. The present invention achieves a balance between the state of charge and the health state of multiple energy storage systems through an improved droop coefficient, thereby effectively extending the service life of the energy storage battery.

[0074] 3. When the power balance changes, the present invention can reasonably adjust the working mode of the energy storage system, realize the autonomous regulation of the power balance of the DC microgrid, stabilize the DC bus voltage, and realize the coordinated control of the photovoltaic system and the energy storage system. BRIEF DESCRIPTION OF THE DRAWINGS

[0075] Figure 1 It is a structural schematic diagram of the optical storage system of the present invention;

[0076] Figure 2 It is a flow chart of the manta ray foraging optimization algorithm combined with the improved adaptive variable step-size conductance increment method in the present invention;

[0077] Figure 3 is ΔD in the present invention 1 Change curve;

[0078] Figure 4 is ΔD in the present invention 2 Change curve;

[0079] Figure 5 is the voltage at the maximum power point of the fixed illumination in the present invention;

[0080] Figure 6This is a comparison diagram of the fixed illumination MPPT effect in the present invention;

[0081] Figure 7 This is the MPPT effect diagram under partial shading in the present invention;

[0082] Figure 8 It is the SOC balance curve diagram in the present invention;

[0083] Fig. 9 This is the SOH equilibrium curve diagram in the present invention. DETAILED DESCRIPTION

[0084] The technical solutions in the embodiments of the present invention will be described clearly and completely below in conjunction with the accompanying drawings, so that those skilled in the art can better understand the advantages and features of the present invention, thereby making a clearer definition of the protection scope of the present invention. The embodiments described in the present invention are only a part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments in the present invention, all other embodiments obtained by ordinary technicians in the field without making creative work are within the scope of protection of the present invention.

[0085] Figure 1 As shown in the structure diagram of the photovoltaic storage system, the photovoltaic system is connected to the DC bus through a DC / DC converter, and the energy storage system is connected to the DC bus through a bidirectional DC / DC converter. A photovoltaic storage system collaborative intelligent control method considering local shading conditions includes the following steps:

[0086] 1) Figure 2 The flowchart of the manta ray foraging optimization algorithm combined with the improved adaptive variable step-size conductance increment method is shown. The optimal solution obtained by the manta ray foraging optimization algorithm is used as the initial duty cycle of the improved adaptive variable step-size conductance increment method.

[0087] When the photovoltaic array is in partial shadow, its output characteristic curve will have multiple peak points. The traditional maximum power point tracking technology is prone to fall into the local optimum, resulting in tracking failure and a significant reduction in the output efficiency of the photovoltaic system.

[0088] Therefore, based on the traditional variable-step-size conductance increment method, it was improved into an adaptive variable-step-size conductance increment method, and the “manta ray foraging optimization algorithm” was introduced to combine it.

[0089] Among them, the manta ray foraging optimization algorithm (MRFO) simulates the foraging process of manta rays in the ocean and mathematically describes the way in which the positions of individual manta rays are updated, thereby realizing the search for the optimal solution in the complex solution space. MRFO can be described as three foraging behaviors, including chain foraging, spiral foraging, and tumbling foraging.

[0090] 1.1) During the chain predation process, the moving direction and step length of the next position of the manta ray individual are determined by the current optimal solution and the previous individual position. The mathematical model of this position update method is as follows:

[0091]

[0092] In the formula, represents the position of the ith individual in the t+1th generation in the d dimension; Respectively represent the positions of the t-th generation, the i-th and the i-1-th individuals in the d dimension; r is a random number uniformly distributed on [0,1]; a is a factor related to r; represents the position of the best individual of the tth generation in the dth dimension; N represents the number of individuals. Represents the distance between the tth generation, the i-th individual and the optimal position in the d-dimension.

[0093] When the MRFO algorithm is used to track the maximum power point of the photovoltaic system, the output power of the photovoltaic array is used as the objective function (fitness), the duty cycle corresponding to the photovoltaic array voltage is used as the individual position, and the duty cycle of the voltage corresponding to the global maximum power point of the photovoltaic array is used as the optimal solution for the population.

[0094] 1.2) When a manta ray finds a prey, due to the existence of chain predation, it is also affected by the previous individual in the process of moving to the current spiral. The mathematical model of this position update method is as follows:

[0095]

[0096] Where T is the total number of iterations; r 1 is a random number uniformly distributed on [0,1]; β is 1 The correlation factor; r is the random number used in step 1.1).

[0097] When t / T≤rand, rand is a random number uniformly distributed in the interval [0,1]. The mathematical equation describing the spiral motion of manta rays can be defined as:

[0098]

[0099] In the formula, represents the random position of the tth generation and the dth dimension; Ub d Indicates the upper bound of the variable value; Lb d represents the lower bound of the variable value; r is the random number used in step 1.1), Ub d -Lb d Indicates the size of the variable interval.

[0100] 1.3) In the rolling predation, the manta ray individual uses the current optimal solution as the rolling fulcrum and rolls to the other side that is a mirror image of its current position. Its mathematical model is expressed as follows:

[0101]

[0102] In the formula, r 2 、r 3 They are all random numbers uniformly distributed in the interval [0,1].

[0103] 1.4) On this basis, the standard deviation of all individuals and the distance between individuals and the optimal individual are introduced into the inertia weight of each individual to judge the convergence speed of the individual. The formula is as follows:

[0104]

[0105] In the formula, ω 0 is a constant, indicating the fixed fluctuation degree of algorithm convergence; ω i Indicates the convergence weight of the i-th individual in the current dimension; λ 1 and λ 2 are all constants; σ represents the standard deviation of all contemporary individual positions in the current dimension; Indicates the distance that the i-th individual in the t-th generation deviates from the current optimal individual position in the d-th dimension.

[0106] Each individual is constantly iterating and updating its speed and position, gradually moving closer to the optimal position. Therefore, when many individuals reach a certain extreme point, it can be considered that the extreme point is the global maximum power point. At this time, σ will be a constant close to 0.

[0107] Therefore, when the individual reaches the vicinity of the global maximum point, it is considered that |ω i -ω 0 | is a minimum value close to 0, at which point |ω i -ω 0 |≤λ 1 a 1 +λ 2 a 2 In the formula, a 1 and a 2 They are all constants close to 0.

[0108] When the external environment changes, the global maximum power point of the photovoltaic system will change, so the algorithm needs to re-track the maximum power point and set the algorithm restart condition to Where P 1 represents the power after the environment changes, and P represents the power before the environment changes.

[0109] 1.5) The process of manta ray foraging optimization algorithm is as follows:

[0110] 1. Set relevant parameters and initialize the population;

[0111] 2. Calculate the initial fitness value;

[0112] 3. Determine whether the condition rand < 0.5 is met. If so, perform spiral foraging. If not, perform chain foraging.

[0113] 4. Calculate the fitness value and update the optimal position;

[0114] 5. Perform tumbling foraging and update position;

[0115] 6. Calculate the fitness value and update the optimal position;

[0116] 7. Determine whether the end condition is met. If so, output the duty cycle corresponding to the maximum power point. Otherwise, repeat steps 2 to 7 to continue iterative updating.

[0117] 2) The traditional conductance increment method determines whether to add or subtract the duty cycle based on the size of dP / dU. When the photovoltaic cell operates at the maximum power point, dP / dU=0; when it is on the left side of the maximum power point, dP / dU>0; when it is on the right side of the maximum power point, dP / dU<0.

[0118] 2.1) Figure 3 ΔD 1 Change curve, ΔD 1 About The increasing function of When ΔD is small, 1 The slope is large when When ΔD is large, 1 The slope is small. That is, when the operating voltage of the photovoltaic array is close to the maximum power point voltage, ΔD 1 Small, can effectively reduce the fluctuation near the maximum power point, improve the accuracy of the algorithm, and its derivative ΔD 1 ' is larger, so it can improve the ΔD 1 When the operating voltage of the photovoltaic array is far away from the maximum power point voltage, ΔD 1 is larger, thus speeding up the dynamic response of the algorithm, and its derivative ΔD 1 ' is small, which can ensure that the algorithm can be adjusted quickly in a short time. Here, the adaptive change formula of step size is proposed:

[0119] dU=U K -U K-1

[0120] dI=I K -I K-1

[0121] P=U K I K

[0122]

[0123] In the formula, I k-1 Represents the current value at the previous moment; I k Represents the current value at the current moment; U k-1 Represents the voltage value at the previous moment; U k represents the voltage value at the current moment; dU represents the change in voltage; dI represents the change in current; P is the power at the current moment; a is about Factor of ΔD 1 Represents each time The step size for resizing; k 1 and c 1 is a constant used to and ΔD 1 At the same time, from Figure 3 It can be seen that we need to select a suitable size of k 1 and c 1 Its amplitude range is limited to meet the appropriate step size change.

[0124] 2.2) Figure 4 ΔD 2 Change curve, in the conductivity increment method, when dU = 0, it is necessary to further determine the size of dI, ΔD 2 is an increasing function of dI. When dI is small, ΔD 2 The slope is large, when dI is large, ΔD 2 The slope is small. That is, when dI is small, it is considered to be close to the maximum power point, ΔD 2 Small, can effectively reduce the fluctuation near the maximum power point, improve the accuracy of the algorithm, and its derivative ΔD 2 ' is larger, so it can improve the ΔD 2 When dI is large, it is considered that it is far away from the maximum power point, ΔD 2 is larger, thus speeding up the dynamic response of the algorithm, and its derivative ΔD 2 ' is relatively small, which can ensure that the algorithm can be adjusted quickly in a short time.

[0125] Because the size and change speed of dI are different from dP / dU, the above formula is not applicable. To address this shortcoming, the above formula is modified and ΔD is proposed. 2 The formula for dI is:

[0126] b=k 2 ·dI

[0127]

[0128] Where b is a factor about dI; ΔD 2 represents the step size adjusted according to the size of dI; k 2 and c 2 are all constants. Figure 4 It can be seen that we need to select a suitable size of k 2 and c 2 Its amplitude range is limited to meet the appropriate step size change.

[0129] 2.3) Figure 5 , Figure 6 The maximum power point voltage and maximum power tracking effect of the improved adaptive variable step size conductance increment method and the traditional conductance increment method under constant light conditions are shown in the figure. A photovoltaic array consisting of five photovoltaic panels connected in series is used, and the light intensity is 1000W / m 2 , the maximum power point voltage of the photovoltaic array is 290V, and the maximum power is 5000W. The solid line is the improved adaptive variable step size conductance increment method, and the dotted line is the traditional conductance increment method. The improved adaptive variable step size conductance increment method completes the maximum power tracking in about 0.2s, and its maximum power point voltage deviation is within 0.1v, and the maximum power error is within 60W; the traditional conductance increment method completes the maximum power tracking in about 0.3s, and its maximum power point voltage deviation is within 0.3v, and the maximum power error is within 150W. It can be seen that the improved maximum power tracking method has higher accuracy and faster adjustment speed.

[0130] 2.4) Figure 7 The maximum power tracking effect under partial shading is shown in the figure. When the light intensity changes, the maximum power point changes accordingly and the maximum power tracking can be completed within 0.1s. It can be seen that it has a good dynamic response effect. The optimal position obtained by the manta ray foraging optimization algorithm is used as the initial duty cycle of the improved adaptive variable step conductance increment method, and then the maximum power tracking is performed. If the algorithm restart condition mentioned in step 1.4) is met, it means that the external light conditions have changed, that is, the maximum power point of the photovoltaic array has changed, so the manta ray foraging optimization algorithm is returned to search for the best result.

[0131] 3) Since the DC / DC converter of the photovoltaic system will cause certain disturbances to the DC bus voltage, the energy storage battery is used to smooth the DC bus voltage disturbance, and the following formula is added to the energy storage battery droop control function:

[0132]

[0133] Among them, U in0Represents the input voltage of the photovoltaic cell DC / DC converter at the last moment, U in1 Represents the input voltage of the photovoltaic cell DC / DC converter at the current moment, D 0 Represents the duty cycle of the photovoltaic cell DC / DC converter at the last moment, D 1 Represents the duty cycle of the photovoltaic cell DC / DC converter at the current moment, Represents the no-load DC reference voltage, ΔU out Represents the disturbance caused by the photovoltaic cell DC / DC converter to the DC bus voltage, U N Represents the nominal voltage of the DC bus.

[0134] 3.1) The droop control function expression is: In the formula, is the reference value of the output voltage of the droop control of the i-th energy storage module, P i is the output power of the i-th energy storage module, R i is the droop coefficient of the i-th energy storage module.

[0135] Taking the dual energy storage in parallel as an example, the droop control function expression is: Since the voltage drop caused by droop control is low, it can be considered that Combined with the droop control expression, the output power ratio between energy storages can be obtained: Where P 1 Represents the output power of the first energy storage battery, P 2 Represents the output power of the second energy storage battery, R 1 Represents the droop factor of the first energy storage battery, R 2 represents the droop factor of the second energy storage battery;

[0136] 3.2) In a system with multiple energy storage systems in parallel, the improved droop coefficient R of the converter of the i-th energy storage system i The calculation formula is as follows:

[0137]

[0138] Among them, R i is the droop coefficient of the i-th energy storage system converter; R 0 is the initial droop coefficient; SOC i represents the i-th energy storage battery; Represents the average state of charge of k energy storage batteries; ΔSOC i represents the difference between the state of charge of the i-th energy storage battery and the average state of charge; Represents the average health status of k energy storage batteries; ΔSOH i Represents the difference between the health status of the i-th energy storage battery and the average health status; β1 , α 1 , β 2 , α 2 are all constants; n is the energy storage module equilibrium speed adjustment factor.

[0139] 3.3) Order BATT i Represents the state of the i-th energy storage battery.

[0140]

[0141] When the energy storage battery works in discharge mode, the ratio of the output power between the energy storage batteries is:

[0142]

[0143] When the energy storage battery works in charging mode, the ratio of the output power between the energy storage batteries is:

[0144]

[0145] Therefore, the output power of the two energy storage batteries can be expressed as:

[0146]

[0147] Among them, P sum Represents the total output power of the two energy storage batteries.

[0148] 3.4) Figure 8 and Fig. 9 As shown in the SOC equilibrium curve and the SOH equilibrium curve, the initial state of charge of the energy storage battery 1 is 49%, and the initial state of health is 92%; the initial state of charge of the energy storage battery 2 is 48.8%, and the initial state of health is 90.5%.

[0149] The solid line in the figure is the state of energy storage battery 1, and the dotted line is the state of energy storage battery 2. In the same time, the SOC and SOH changes of energy storage battery 1 are greater than those of energy storage battery 2, and the SOC and SOH of the two energy storage batteries tend to be unified in about 25s. When the energy storage module is in the discharge mode, the energy storage battery with larger SOC and SOH has a smaller adaptive droop coefficient, the converter discharge current is larger, and the output power is larger; the energy storage battery with smaller SOC and SOH has a larger adaptive droop coefficient, the converter discharge current is smaller, and the output power is smaller, so that the battery state reaches the same level. When the energy storage module is in the charging mode, the energy storage battery with larger SOC and SOH has a larger adaptive droop coefficient, the converter discharge current is smaller, and the output power is smaller; the energy storage battery with smaller SOC and SOH has a smaller adaptive droop coefficient, and its converter discharge current is larger and the output power is larger. Through the above, the SOC balance of the energy storage module is achieved, and the discharge depth of the energy storage module is adjusted at the same time, so as to achieve SOH balance.

[0150] In summary, the present invention combines the manta ray foraging optimization algorithm with the improved adaptive variable step-size conductance increment method to achieve maximum power tracking of partially shaded photovoltaic arrays, thereby improving energy utilization efficiency; through the improved droop coefficient, the charge state and health state of multiple energy storage systems are balanced, effectively extending the service life of the energy storage battery; when the power balance changes, the present invention can reasonably adjust the working mode of the energy storage system, realize autonomous regulation of the power balance of the DC microgrid, stabilize the DC bus voltage, and realize coordinated control of the photovoltaic system and the energy storage system.

[0151] The description and practice disclosed in the present invention are easy to think and understand for ordinary technicians in the technical field, and several improvements and modifications can be made without departing from the principles of the present invention. Therefore, modifications or improvements made without departing from the spirit of the present invention should also be regarded as the protection scope of the present invention.

Claims

1. A method for cooperative intelligent control of a photovoltaic storage system taking into account local shading conditions, characterized in that: The steps include: The traditional variable-step conductance increment method is improved into an adaptive variable-step conductance increment method, and the manta ray foraging optimization algorithm MRFO is introduced to combine it; among them, the manta ray foraging optimization algorithm has three foraging behaviors, including chain foraging, spiral foraging and tumbling foraging: Step 1.1) During the chain predation process, the moving direction and step length of the next position of the manta ray individual are determined by the current optimal solution and the previous individual position. The mathematical model of this position update method is as follows: In the formula, represents the position of the ith individual in the t+1th generation in the d dimension; Respectively represent the positions of the t-th generation, the i-th and the i-1-th individuals in the d dimension; r is a random number uniformly distributed on [0,1]; a is a factor related to r; represents the position of the best individual of the tth generation in the dth dimension; N represents the number of individuals, Indicates the distance between the t-th generation, the i-th individual and the optimal position in the dimension in d dimension; When the MRFO algorithm is used to track the maximum power point of the photovoltaic system, the output power of the photovoltaic array is used as the objective function, the duty cycle corresponding to the photovoltaic array voltage is used as the individual position, and the duty cycle of the voltage corresponding to the global maximum power point of the photovoltaic array is used as the optimal solution of the population; Step 1.2) When a manta ray finds a prey, due to the existence of the chain predation method, the manta ray is also affected by the previous individual in the process of moving to the current spiral. The mathematical model of this position update method is as follows: Where T is the total number of iterations; r1 is a random number uniformly distributed on [0,1]; β is a factor related to r1; r is the random number used in step 1.1); When t / T≤rand, rand is a random number uniformly distributed in the interval [0,1]. The mathematical equation describing the spiral motion of manta rays is defined as: In the formula, represents the random position of the tth generation and the dth dimension; Ub d Indicates the upper bound of the variable value; Lb d represents the lower bound of the variable value; r is the random number used in step 1.1), Ub d -Lb d Indicates the size of the variable interval; Step 1.3) In the rolling predation, the manta ray uses the current optimal solution as the rolling fulcrum and rolls to the other side that is a mirror image of its current position. The mathematical model is expressed as follows: In the formula, r2 and r3 are random numbers uniformly distributed in the interval [0,1]; Step 1.4) On this basis, the standard deviation of all individuals and the distance between individuals and the optimal individual are introduced into the inertia weight of each individual to judge the convergence speed of the individual. The formula is as follows: Where ω0 is a constant, indicating the fixed fluctuation degree of algorithm convergence; ω i represents the convergence weight of the i-th individual in the current dimension; λ1 and λ2 are both constants; σ represents the standard deviation of all contemporary individual positions in the current dimension; Indicates the distance of the ith individual in the tth generation from the current optimal individual position in the dth dimension; Each individual is constantly iterating and updating its speed and position, gradually moving closer to the optimal position. Therefore, when many individuals reach a certain extreme point, it can be considered that the extreme point is the global maximum power point. At this time, σ will be a constant close to 0. Therefore, when the individual reaches the vicinity of the global maximum point, it is considered that |ω i -ω0| is a minimum value close to 0, at which point |ω i -ω0|≤λ1a1+λ2a2; where a1 and a2 are constants close to 0; When the external environment changes, the global maximum power point of the photovoltaic system will change, so the algorithm needs to re-track the maximum power point and set the algorithm restart condition to In the formula, P1 represents the power after the environment changes, and P represents the power before the environment changes; Step 1.5) The process of manta ray foraging optimization algorithm is as follows: Step 1.

51. Set relevant parameters and initialize the population; Step 1.

52. Calculate the initial fitness value; Step 1.

53. Determine whether the condition rand < 0.5 is met. If so, perform spiral foraging. If not, perform chain foraging. Step 1.

54. Calculate the fitness value and update the optimal position; Step 1.

55. Perform tumbling foraging and update the position; Step 1.

56. Calculate the fitness value and update the optimal position; Step 1.

57. Determine whether the end condition is met. If so, output the duty cycle corresponding to the maximum power point. Otherwise, repeat steps 1.52 to 1.57 to continue iterative updating.

2. The method for cooperative intelligent control of a photovoltaic storage system taking into account local shading conditions according to claim 1, characterized in that: The traditional conductivity increment method determines whether to add or subtract the duty cycle based on the size of dP / dU. When the photovoltaic cell is working at the maximum power point, dP / dU = 0; when it is on the left side of the maximum power point, dP / dU > 0; when it is on the right side of the maximum power point, dP / dU < 0; Step 2.1) Adaptive variable step size algorithm On this basis, it is necessary to ensure that the step size ΔD1 is a monotonically increasing function of dP / dU. Here, the adaptive change formula of the step size is proposed: dU=U K -U K-1 dI=I K -I K-1 P=U K ·AND K In the formula, I k-1 Represents the current value at the previous moment; I k Represents the current value at the current moment; U k-1 Represents the voltage value at the previous moment; U k Represents the voltage value at the current moment; dU represents the change in voltage; dI represents the change in current; P is the power at the current moment; a is about factor; ΔD1 represents each time according to The step size of the size adjustment; k1 and c1 are constants used to adjust Scale with ΔD1 to make its size within a suitable range; Step 2.2) In the conductance increment method, when dU = 0, it is necessary to further determine the size of dI; Because the size and change speed of dI are different from dP / dU, the above formula is not applicable. To address this shortcoming, the above formula is modified and the formula of ΔD2 with respect to dI is proposed: b=k2·dI Where b is a factor related to dI; ΔD2 represents the step size adjusted according to the size of dI; k2 and c2 are both constants; In step 2.3), the optimal position obtained by the manta ray foraging optimization algorithm is used as the initial duty cycle of the improved adaptive variable step size conductance increment method, and then the maximum power tracking is performed; if the algorithm restart condition mentioned in step 1.4) is met, it means that the external lighting conditions have changed, that is, the maximum power point of the photovoltaic array has changed, so the manta ray foraging optimization algorithm is returned to search for the best result.

3. The method for cooperative intelligent control of a photovoltaic storage system taking into account local shading conditions according to claim 1, characterized in that: Since the photovoltaic system DC / DC converter brings certain disturbances to the DC bus voltage, energy storage batteries are used to smooth the DC bus voltage disturbances. The following formula is added to the energy storage battery droop control function: Among them, U in0 Represents the input voltage of the photovoltaic cell DC / DC converter at the last moment, U in1 represents the input voltage of the photovoltaic cell DC / DC converter at the current moment, D0 represents the duty cycle of the photovoltaic cell DC / DC converter at the previous moment, and D1 represents the duty cycle of the photovoltaic cell DC / DC converter at the current moment. Represents the no-load DC reference voltage, ΔU out Represents the disturbance caused by the photovoltaic cell DC / DC converter to the DC bus voltage, U N Represents the nominal voltage of the DC bus; Step 3.1) The droop control function expression is: In the formula, is the reference value of the output voltage of the droop control of the i-th energy storage module, P i is the output power of the i-th energy storage module, R i is the droop coefficient of the i-th energy storage module; Taking the dual energy storage in parallel as an example, the droop control function expression is: Since the voltage drop caused by droop control is low, it is considered Combined with the droop control expression, the output power ratio between energy storages is obtained Where P1 represents the output power of the first energy storage battery, P2 represents the output power of the second energy storage battery, R1 represents the droop coefficient of the first energy storage battery, and R2 represents the droop coefficient of the second energy storage battery; Step 3.2) In a system with multiple energy storages in parallel, the improved droop coefficient R of the converter of the i-th energy storage system i The calculation formula is as follows: Among them, R i is the droop coefficient of the i-th energy storage system converter; R0 is the initial droop coefficient; SOC i represents the i-th energy storage battery; Represents the average state of charge of k energy storage batteries; ΔSOC i represents the difference between the state of charge of the i-th energy storage battery and the average state of charge; Represents the average health status of k energy storage batteries; ΔSOH i It represents the difference between the health status of the i-th energy storage battery and the average health status; β1, α1, β2, α2 are all constants; n is the energy storage module balancing speed adjustment factor; Step 3.3) Order BATT i Indicates the state of the i-th energy storage battery; When the energy storage battery works in discharge mode, the ratio of the output power between the energy storage batteries is: When the energy storage battery works in charging mode, the ratio of the output power between the energy storage batteries is: Therefore, the output power of the two energy storage batteries is expressed as: Among them, P sum Represents the total output power of the two energy storage batteries; When the energy storage module is in the discharge mode, the energy storage battery with larger SOC and SOH has a smaller adaptive droop coefficient, the converter discharge current is larger, and the output power is larger; the energy storage battery with smaller SOC and SOH has a larger adaptive droop coefficient, the converter discharge current is smaller, and the output power is smaller, so that the battery status reaches the same level; when the energy storage module is in the charging mode, the energy storage battery with larger SOC and SOH has a larger adaptive droop coefficient, the converter discharge current is smaller, and the output power is smaller; the energy storage battery with smaller SOC and SOH has a smaller adaptive droop coefficient, and its converter discharge current is larger and the output power is larger.

Images