阅读说明：本技术 一种基于樽海鞘算法的仿生智能孤岛检测方法 (Bionic intelligent island detection method based on goblet sea squirt algorithm ) 是由 胡丹丹 赵倩 于 2019-09-19 设计创作，主要内容包括：本发明涉及一种基于樽海鞘算法的仿生智能孤岛检测方法,它包括以下步骤：S1：采用改进的樽海鞘算法进行实时最大功率跟踪；S2：利用过/欠压检测法、过/欠频检测法实时检测；S3：周期性地扰动占空比,改变逆变器输出功率与负载功率的匹配程度；S4：检测外部环境是否发生突变；该发明提供一种适用于多变环境下的基于樽海鞘算法的仿生智能孤岛检测方法,提高了检测方法的适应性；周期性地扰动占空比,间接改变逆变器输出功率与负载功率的匹配程度,进而避免检测盲区,提高方法的效率；本发明具有检测效率、适应性和稳定性高的优点。(the invention relates to a bionic intelligent island detection method based on a goblet sea squirt algorithm, which comprises the following steps: s1: performing real-time maximum power tracking by adopting an improved goblet sea squirt algorithm; s2: detecting in real time by using an over/under voltage detection method and an over/under frequency detection method; s3: periodically disturbing the duty ratio and changing the matching degree of the output power of the inverter and the load power; s4: detecting whether mutation occurs in the external environment; the bionic intelligent island detection method based on the goblet and ascidian algorithm is suitable for variable environments, and the adaptability of the detection method is improved; the duty ratio is disturbed periodically, the matching degree of the output power of the inverter and the load power is indirectly changed, the detection blind area is further avoided, and the efficiency of the method is improved; the invention has the advantages of high detection efficiency, high adaptability and high stability.)

1. a bionic intelligent island detection method based on a goblet sea squirt algorithm is characterized by comprising the following steps: it comprises the following steps:

S1: performing real-time maximum power tracking by adopting an improved goblet sea squirt algorithm;

S2: detecting in real time by using an over/under voltage detection method and an over/under frequency detection method;

S3: periodically disturbing the duty ratio and changing the matching degree of the output power of the inverter and the load power;

S4: and detecting whether mutation occurs in the external environment.

2. The bionic intelligent island detection method based on the turtle sea squirt algorithm as claimed in claim 1, wherein: the improved goblet sea squirt algorithm in step S1 is composed of a leader and a follower, wherein the leader is updated according to equation (1):

Wherein the content of the first and second substances,Representing the position of the first leader in the j-th dimension; f_jRepresenting the location of the food source in dimension j; ub_j,ib_jRepresenting the upper and lower boundaries of the j-th dimension search space; c. C₁、c₂、c₃Is a random number, wherein c₂、c₃in the range of [0,1]having the effect of enhancing the randomness of the movement of the leader and enhancing the search capability of the whole world, c₁is the most important parameter, and can be regarded as a decreasing function from 2 to 0, and the expression is shown in formula (2):

Wherein l is the current iteration number, l_maxis the maximum number of iterations, c₁Is often called convergence factor and is used for balancing the exploration and development capability of the algorithm in the iterative process, when the convergence factor is more than 1, the algorithm carries out global exploration, and when the convergence factor is more than 1When the value is less than 1, the algorithm is locally developed, then the optimal value is accurately searched, the follower moves along with the leader in a chain shape, and the position is updated according to the formula (3):

3. the bionic intelligent island detection method based on the turtle sea squirt algorithm as claimed in claim 2, wherein: when the position of the follower is updated, the improved goblet sea squirt algorithm increases the inertia weight w which is linearly reduced, is used for accelerating the convergence speed of the algorithm, and is updated according to the formula (4):

where l is the current iteration number, iter_maxis the maximum number of iterations.

4. The bionic intelligent island detection method based on the turtle sea squirt algorithm as claimed in claim 1, wherein: the judgment basis of the over/under voltage detection method and the over/under frequency detection method in the step S2 is as follows: if the voltage of the grid-connected point U_PCC1The amplitude and the frequency of the inverter meet any one of conditions a, b and c, namely, the island state can be determined according to an over/under voltage method and an over/under frequency method, and when an island effect is detected, the power tube of the inverter is immediately disconnected, so that a local load is isolated;

a：U_PCC1<0.88U_N；

b：U_PCC1>1.1U_N；

c：0.5Hz<|f-f_g|；

Wherein, U_PCC1The grid-connected point voltage after the island is formed; u shape_NRated voltage for the power grid; f is the grid-connected point voltage frequency, f_gIs the grid voltage frequency.

5. The bionic intelligent island detection method based on the turtle sea squirt algorithm as claimed in claim 1, wherein: in step S3, duty cycle perturbation is periodically performed according to equation (5):

wherein D is₀、D₁The duty ratios of the Boost circuit before and after disturbance are respectively obtained, the disturbance period is 1s, the disturbance duration of each period is 2 power frequency periods, and the matching degree of the output power of the inverter and the load power is indirectly changed by properly disturbing the duty ratio D, so that the island effect is detected.

6. The bionic intelligent island detection method based on the turtle sea squirt algorithm as claimed in claim 1, wherein: the basis for judging the sudden change of the external environment in step S4 is as shown in formula (6):

Wherein, P_realis the actual output power of the photovoltaic array; p_mMaximum output power of the photovoltaic array, Δ P is the threshold for photovoltaic array variation, set herein to 0.5.

7. The bionic intelligent island detection method based on the turtle sea squirt algorithm as claimed in claim 2, wherein: the improved goblet ascidian algorithm comprises the following specific steps:

S11: initializing individual positions of a population;

S12: calculating fitness values, arranging individuals according to the fitness values, selecting the first individual as a leader and the other individuals as followers, and recording the position of the first individual as an optimal food source;

S13: updating the positions of the leader and the follower respectively according to the formula (1) and the formula (3);

s14: calculating a fitness value and updating the position of the food source;

S15: repeating S12, S13 and S14 until the algorithm termination condition is met;

s16: and (4) reinitializing, and when the formula (6) is met, reinitializing the algorithm.

Technical Field

the invention relates to the technical field of fault detection of photovoltaic power generation systems, in particular to a bionic intelligent island detection method based on a goblet sea squirt algorithm.

Background

With the rapid development of global renewable energy, the distributed power generation system generates power by using local renewable energy, and has the characteristics of economy, high efficiency, environmental protection, cleanness and the like, so that the distributed power generation system is widely researched and developed, and brings new problems, and one of the key problems is how to detect the islanding effect.

the islanding effect means that power supply of a power grid is interrupted due to reasons such as electrical faults or natural factors, a photovoltaic grid-connected power generation system does not detect a power failure state and continues to supply power to surrounding loads, so that a self-supply system which is separated from the control of a power grid company is formed, and serious consequences such as damage to power utilization equipment, threat to personal safety of maintainers, reclosing failure and the like can be brought about, so that the islanding effect detection method has important significance for research.

The island detection method mainly comprises a sliding mode frequency drift method, an active frequency shift method, an active current disturbance method and the like, wherein the method effectively reduces a detection blind area but influences the quality of electric energy, the passive method mainly focuses on output parameters of an inverter, does not need to inject disturbance quantity, does not influence the quality of electric energy and the stability of a system, but has a large detection blind area which cannot reliably and stably detect the island effect, and meanwhile, the method does not consider the influence of an external environment on the detection method; therefore, it is very necessary to provide a bionic intelligent island detection method based on the goblet sea squirt algorithm with high detection efficiency, adaptability and stability.

Disclosure of Invention

the invention aims to overcome the defects of the prior art and provides a bionic intelligent island detection method based on a goblet sea squirt algorithm, which has high detection efficiency, adaptability and stability.

The purpose of the invention is realized as follows: a bionic intelligent island detection method based on a goblet sea squirt algorithm comprises the following steps:

S1: performing real-time maximum power tracking by adopting an improved goblet sea squirt algorithm;

S2: detecting in real time by using an over/under voltage detection method and an over/under frequency detection method;

S3: periodically disturbing the duty ratio and changing the matching degree of the output power of the inverter and the load power;

S4: and detecting whether mutation occurs in the external environment.

The improved goblet sea squirt algorithm in the step S1 is composed of a leader and a follower, and the leader is updated according to the formula (1):

wherein the content of the first and second substances,Representing the position of the first leader in the j-th dimension; f_jRepresenting the location of the food source in dimension j; ub_j,ib_jRepresenting the upper and lower boundaries of the j-th dimension search space; c. C₁、c₂、c₃is a random number, wherein c₂、c₃In the range of [0,1]Having the effect of enhancing the randomness of the movement of the leader and enhancing the search capability of the whole world, c₁is the most important parameter, and can be regarded as a decreasing function from 2 to 0, and the expression is shown in formula (2):

wherein l is the current iteration number, l_maxIs the maximum number of iterations, c₁The method is often called as a convergence factor and is used for balancing exploration and development capacity of the algorithm in an iterative process, when the convergence factor is larger than 1, the algorithm carries out global exploration, when the convergence factor is smaller than 1, the algorithm carries out local development, further an optimal value is accurately found, a follower moves along with the leader in a chain shape, and the position is updated according to an equation (3):

when the position of the follower is updated, the improved goblet sea squirt algorithm increases the inertia weight w which is linearly reduced, is used for accelerating the convergence speed of the algorithm, and is updated according to the formula (4):

Where l is the current iteration number, iter_maxIs the maximum number of iterations.

the judgment basis of the over/under voltage detection method and the over/under frequency detection method in the step S2 is as follows: if the voltage of the grid-connected point U_PCC1The amplitude and the frequency of the inverter meet any one of conditions a, b and c, namely, the island state can be determined according to an over/under voltage method and an over/under frequency method, and when an island effect is detected, the power tube of the inverter is immediately disconnected, so that a local load is isolated;

a：U_PCC1<0.88U_N；

b：U_PCC1>1.1U_N；

c：0.5Hz<|f-f_g|；

Wherein, U_PCC1The grid-connected point voltage after the island is formed; u shape_NRated voltage for the power grid; f is the grid-connected point voltage frequency, f_gis the grid voltage frequency.

In step S3, duty cycle perturbation is periodically performed according to equation (5):

Wherein D is₀、D₁The duty ratios of the Boost circuit before and after disturbance are respectively obtained, the disturbance period is 1s, the disturbance duration of each period is 2 power frequency periods, and the matching degree of the output power of the inverter and the load power is indirectly changed by properly disturbing the duty ratio D, so that the island effect is detected.

The basis for judging the sudden change of the external environment in step S4 is as shown in formula (6):

Wherein, P_realis the actual output power of the photovoltaic array; p_mMaximum output power of the photovoltaic array, Δ P is the threshold for photovoltaic array variation, set herein to 0.5.

The improved goblet ascidian algorithm comprises the following specific steps:

S11: initializing individual positions of a population;

S12: calculating fitness values, arranging individuals according to the fitness values, selecting the first individual as a leader and the other individuals as followers, and recording the position of the first individual as an optimal food source;

s13: updating the positions of the leader and the follower respectively according to the formula (1) and the formula (3);

s14: calculating a fitness value and updating the position of the food source;

S15: repeating S12, S13 and S14 until the algorithm termination condition is met;

S16: and (4) reinitializing, and when the formula (6) is met, reinitializing the algorithm.

the invention has the beneficial effects that: firstly, a bionic intelligent island detection method based on a goblet and sea squirt algorithm and suitable for a changeable environment is provided, and the adaptability of the algorithm is improved; secondly, the duty ratio is disturbed periodically, the matching degree of the output power of the inverter and the load power is indirectly changed, the detection blind area is further avoided, and the efficiency of the method is improved; the invention has the advantages of high detection efficiency, high adaptability and high stability.

Drawings

Fig. 1 is a schematic structural diagram of a commonly-used two-stage photovoltaic grid-connected system.

FIG. 2 is a flow chart of a bionic intelligent island detection method based on the goblet sea squirt algorithm.

FIG. 3 is a normal disturbance waveform diagram under uniform illumination of a photovoltaic grid-connected system.

FIG. 4 shows P before and after disturbance of uniform illumination and island state of a photovoltaic grid-connected system_PV、D、U_PV、U_bus、U_PCCa、I_PCCaAnd (4) waveform diagrams.

fig. 5 is a voltage and current waveform diagram of grid-connected points before and after uniform illumination and island state disturbance of a photovoltaic grid-connected system.

FIG. 6 shows the effective value of grid-connected point voltage before and after the disturbance of the uniform illumination and island state of the grid-connected photovoltaic system and the sum of the effective value of grid-connected point voltage and the sum of U_NIs shown in the ratio waveform.

FIG. 7 shows P before and after disturbance in island state when output power of inverter is not matched with load power_PV、D、U_PV、U_bus、U_PCCa、I_PCCaand (4) waveform diagrams.

Fig. 8 is a waveform diagram of effective values of grid-connected point voltages before and after the disturbance of the uniform illumination and the island state of the photovoltaic grid-connected system and the ratio of the effective values to the rated voltage.

FIG. 9 shows P before and after disturbance in a state of uneven illumination and island of a photovoltaic grid-connected system_PV、D、U_PV、U_bus、U_PCCa、I_PCCaand (4) waveform diagrams.

Fig. 10 is a voltage and current waveform diagram of grid-connected points before and after disturbance in a photovoltaic grid-connected system under uneven illumination and island state.

FIG. 11 shows effective values U of grid-connected point voltages before and after disturbance in an island state due to uneven illumination of a photovoltaic grid-connected system_{a_rms}And U_NIs shown in the ratio waveform.

FIG. 12 shows a U value under uneven illumination of a photovoltaic grid-connected system_PCC＝1.1U_Nand before and after disturbance in island state_PV、D、U_PV、U_bus、U_PCCa、I_PCCaAnd (4) waveform diagrams.

FIG. 13 shows a U under uneven illumination of a photovoltaic grid-connected system_PCC＝1.1U_Nand voltage and current oscillograms of grid-connected points before and after disturbance in an island state.

FIG. 14 shows a U value under uneven illumination of a photovoltaic grid-connected system_PCC＝1.1U_NAnd the effective value U of the grid-connected point voltage before and after disturbance in the island state_{a_rms}And U_NIs shown in the ratio waveform.

Detailed Description

the invention is further described below with reference to the accompanying drawings.