Method for evaluating power supply adequacy of solar power supply system in cloudy and rainy days

Abstract

The invention discloses a method for evaluating the power supply adequacy of a solar power supply system in cloudy and rainy days, belongs to the technical field of power supply, and can monitor the running state of the solar power supply system in real time and predict the electrical load of a solar power supply area, thereby accurately evaluating the power supply adequacy under cloudy and rainy weather conditions. A user can conveniently adjust the power demand according to a power supply adequacy result, and energy optimization configuration is realized; the method has the advantages of simple operation, high practicability, wide application range and the like.

Description

Technical Field

[0001] The invention belongs to the technical field of power supply, and in particular relates to a method for evaluating power supply adequacy of a solar power supply system for rainy days. Background Art

[0002] A solar power supply system is a power generation system that uses solar panels to convert sunlight into electrical energy. Solar panels generate direct current under sunlight, which is regulated by a controller. One part is used to directly supply power to the DC load, and the other part is used to charge the battery. When the power generated by the solar panel is not enough to meet the load demand, the battery will discharge and supply power to the load through the controller. If the system is equipped with an inverter, the direct current can be converted into alternating current for use by the AC load, or connected to the public power grid. With the popularization of solar power generation technology, more and more solar power supply systems are used in daily life. However, solar power generation is greatly affected by weather conditions, especially in rainy weather, the power generation capacity of the solar power supply system drops significantly, resulting in unstable power supply. In the prior art, there is a lack of an evaluation method for the power supply adequacy of the solar power supply system under rainy weather conditions, which makes it impossible for users to accurately grasp the power supply situation on rainy days, affecting normal power consumption. Summary of the invention

[0003] The present invention provides a method for evaluating the power supply adequacy of a solar power supply system on rainy days, so as to solve the problem in the prior art that it is difficult to evaluate the power supply adequacy of solar power supply on rainy days.

[0004] A method for evaluating the power supply adequacy of a solar power supply system on rainy days, comprising:

[0005] Collecting the solar power supply influencing factors and daily power supply under rainy days, and using the GISA-BP algorithm to learn the solar power supply influencing factors and daily power supply to determine a first prediction rule;

[0006] Collecting regional power consumption characteristics and daily power consumption corresponding to the solar power supply area, and using the GISA-BP algorithm to learn the regional power consumption characteristics and daily power consumption to determine a second prediction rule;

[0007] Collecting real-time solar power supply influencing factors on rainy days, and analyzing the real-time solar power supply factors using the first prediction rule to determine the predicted solar power supply amount;

[0008] Collecting real-time regional electricity consumption characteristics corresponding to the solar power supply area, and using the second prediction rule to analyze the real-time regional electricity consumption characteristics to determine the predicted regional electricity consumption;

[0009] According to the predicted solar power supply and the predicted regional power consumption, a power supply adequacy assessment is performed on the solar power supply system on rainy days to obtain a power supply adequacy assessment result.

[0010] In a possible implementation, collecting the solar power supply influencing factors and daily power supply under rainy days includes:

[0011] The average temperature, daily temperature range, average relative humidity, average wind speed, sunshine hours, visibility, global horizontal radiation, global oblique radiation, diffuse horizontal radiation and diffuse oblique radiation under rainy and cloudy conditions were collected to obtain the solar power supply impact factor;

[0012] Collect daily power supply under rainy and cloudy conditions.

[0013] In a possible implementation manner, the GISA-BP algorithm is used to learn the solar power supply influencing factor and the daily power supply to determine the first prediction rule, including:

[0014] A BP algorithm is used to create a first initial rule, and the parameters of the first initial rule are initialized to obtain a first parameter individual, and a plurality of different first parameter individuals are repeatedly obtained;

[0015] The solar power supply influencing factors are formed into vectors and used as input data of the first initial rule. The daily power supply corresponding to the solar power supply influencing factors is used as the expected label, and the loss function value corresponding to each first parameter individual is obtained;

[0016] Determine the optimal first parameter individual according to the loss function value corresponding to each first parameter individual;

[0017] Based on the optimal first parameter individual, each first parameter individual is updated by adopting an adaptive vibration-guided search strategy to obtain a first parameter individual after vibration-guided search;

[0018] For the first parameter individuals after the vibration-guided search, an adaptive local area search strategy is adopted to update each first parameter individual to obtain the first parameter individual after the local area search;

[0019] For the first parameter individuals after the local area search, an adaptive population collaborative search strategy is used to update each first parameter individual to obtain the first parameter individuals after the population collaborative search;

[0020] Repeat the vibration-guided search, local area search, and population collaborative search until the number of training times reaches the maximum number of training times, re-obtain the optimal first parameter individual, and apply the parameters contained in the optimal first parameter individual to the first initial rule to obtain the first prediction rule.

[0021] In a possible implementation, based on the optimal first parameter individual, each first parameter individual is updated using an adaptive vibration-guided search strategy to obtain a first parameter individual after vibration-guided search, including:

[0022] Based on the current number of training times, the adaptive attenuation factor is determined as:

[0023]

[0024] Among them, z t represents the adaptive attenuation factor during the t-th training process, z max Indicates the maximum value corresponding to the adaptive attenuation factor, z min represents the minimum value corresponding to the adaptive attenuation factor, e represents the natural constant, and T represents the maximum number of training times;

[0025] According to the adaptive attenuation factor, the first adaptive vibration induction coefficient and the second adaptive vibration induction coefficient are obtained as follows:

[0026] α 1 =2z t r 1 -z t

[0027] α 2 =kz t r 2 +1

[0028] Among them, α 1 represents the first adaptive vibration guidance coefficient, r 1 Represents the first random number between (0, 1), α 2 represents the second adaptive vibration guidance coefficient, k represents the natural constant, r 2 represents a second random number between (0, 1);

[0029] Generate a third random number r between (0, 1) 3 , and determine the third random number r 3 Is it less than the first decision probability? If so, based on the optimal first parameter individual and according to the first adaptive vibration guidance coefficient and the second adaptive vibration guidance, a first vibration guidance search is performed on the first parameter individual to obtain the first parameter individual after the vibration guidance search; otherwise, based on the optimal first parameter individual and according to the first adaptive vibration guidance coefficient and the second adaptive vibration guidance, a second vibration guidance search is performed on the first parameter individual;

[0030] Based on the optimal first parameter individual, and according to the first adaptive vibration guidance coefficient and the second adaptive vibration guidance, the first parameter individual is searched for the first vibration guidance as follows:

[0031]

[0032] in, represents the i-th first parameter individual in the t-th training process, i=1,2,…,I, I represents the total number of first parameter individuals, represents the optimal first parameter individual, Represents the first parameter individual after vibration-guided search

[0033] Based on the optimal first parameter individual, and according to the first adaptive vibration guidance coefficient and the second adaptive vibration guidance, the second vibration guidance search is performed on the first parameter individual as follows:

[0034]

[0035] Among them, π represents pi, cos represents cosine function, and sin represents sine function.

[0036] In a possible implementation, for the first parameter individuals after the vibration-guided search, an adaptive local area search strategy is used to update each first parameter individual to obtain the first parameter individual after the local area search, including:

[0037] Based on the current number of training times, the adaptive local development factor is determined as:

[0038] p=(1-t / T) t / T

[0039] Where p represents the adaptive local development factor;

[0040] For the first parameter individual after the vibration-guided search, generate a fourth random number r between (0, 1) 4 , and determine the fourth random number r 4 Is it less than the second decision probability? If so, perform a first local area search on the first parameter individual to obtain the first parameter individual after the local area search; otherwise, perform a second local area search on the first parameter individual to obtain the first parameter individual after the local area search;

[0041] The first local area search for the first parameter individual is:

[0042]

[0043] in, represents the first parameter individual after the nth vibration-guided search, n = 1, 2, ..., I, r 5 represents the fifth random number between (0, 1), and F represents the random quantity between [-1, 1];

[0044] The second local area search for the first parameter individual is:

[0045]

[0046] in, Represents the first parameter individual after the local area search

[0047] In a possible implementation, for the first parameter individuals after the local area search, an adaptive population collaborative search strategy is used to update each first parameter individual to obtain the first parameter individuals after the population collaborative search, including:

[0048] Based on the current number of training times, the adaptive cooperation factor is determined as:

[0049]

[0050] in, represents the adaptive cooperation factor corresponding to the first parameter individual after the mth local area search in the tth training process, β 0 represents the first preset constant, represents the intermediate coefficient corresponding to the first parameter individual after the mth local area search, m = 1, 2, ..., I, ξ max represents the maximum value corresponding to the intermediate coefficient, ξ max represents the minimum value corresponding to the intermediate coefficient, It represents the fitness value corresponding to the first parameter individual after the mth local area search. represents the minimum fitness value in the first parameter individual after the local area search, represents the average fitness value of the first parameter individual after the local area search; the fitness value is obtained by adding the loss function value and the second preset constant and taking the reciprocal;

[0051] For the first parameter individual after the mth local area search, according to the adaptive cooperation factor, the cooperation influence of all other first parameter individuals on the first parameter individual is determined as follows:

[0052]

[0053] in, represents the collaborative influence, r 6 represents the sixth random number between (0, 1), Represents the first parameter individual after the kth local area search The first parameter individual after the mth local area search Regarding the influence of the d-th dimension parameter, d=1, 2, ..., D, D represents the total dimension of the parameters of the first parameter individual, ε represents the second preset constant, and ε=0.0001; Represents the first parameter individual The corresponding weight, Represents the first parameter individual The corresponding weight, and The method of obtaining is the same; Represents the first parameter individual The d-th dimension parameter of Represents the first parameter individual The d-th dimension parameter of Represents the first parameter individual The corresponding position quality, Represents the first parameter individual The corresponding position quality, f m represents the fitness value corresponding to the first parameter individual after the mth local area search, f w represents the minimum fitness value of the first parameter individual after the local area search, f best Represents the maximum fitness value of the first parameter individual after the local area search;

[0054] Based on collaborative influence And weight Determine the first parameter individual The collaborative influence term corresponding to the d-th dimension parameter is:

[0055] According to the first parameter The collaborative influence term corresponding to the d-th dimension parameter Determine the first parameter individual The incremental term corresponding to the d-th dimension parameter is:

[0056]

[0057] in, Indicates the first parameter individual in the tth training process The incremental term corresponding to the d-th dimension parameter of Indicates the first parameter individual in the t+1th training process The incremental term corresponding to the d-th dimension parameter, r 7 Represents the seventh random number between (0,1), A 1 Represents the first constant between (0,1), r 8 Represents the eighth random number between (-1,1), represents the d-th dimension parameter of the optimal first parameter individual, A2 Represents the second constant between (0,1), r 9 Represents the ninth random number between (-1,1);

[0058] According to the first parameter The incremental term corresponding to the d-th dimension parameter updates the first parameter individual as follows:

[0059]

[0060] in, Represents the first parameter individual after the population collaborative search The d-th dimension parameter of .

[0061] In a possible implementation, collecting regional power consumption characteristics and daily power consumption corresponding to the solar power supply area includes:

[0062] Collect seasonal data, weather data, temperature data and date type data of the power supply area to obtain regional electricity consumption characteristics corresponding to the solar power supply area;

[0063] Collect the daily power consumption corresponding to the power consumption characteristics of the area;

[0064] The season data includes spring, summer, autumn or winter, the weather data includes rainy, sunny, cloudy or snowy, and the date type data includes working days or non-working days.

[0065] In a possible implementation, the GISA-BP algorithm is used to learn the regional power consumption characteristics and the daily power consumption to determine the second prediction rule, including:

[0066] A BP algorithm is used to create a second initial rule, and the parameters of the second initial rule are initialized to obtain a second parameter individual, and a plurality of different second parameter individuals are repeatedly obtained;

[0067] The regional electricity consumption features are formed into vectors and used as input data for the second initial rule. The daily electricity consumption corresponding to the regional electricity consumption features is used as the expected label to obtain the loss function value corresponding to each second parameter individual;

[0068] Determine the optimal second parameter individual according to the loss function value corresponding to each second parameter individual;

[0069] Based on the optimal second parameter individual, each second parameter individual is updated by adopting an adaptive vibration guided search strategy to obtain a second parameter individual after vibration guided search;

[0070] For the second parameter individuals after the vibration-guided search, an adaptive local area search strategy is adopted to update each second parameter individual to obtain the second parameter individuals after the local area search;

[0071] For the second parameter individuals after the local area search, an adaptive population collaborative search strategy is used to update each second parameter individual to obtain the second parameter individuals after the population collaborative search;

[0072] Repeat the vibration-guided search, local area search and population collaborative search until the number of training times reaches the maximum number of training times, re-obtain the optimal second parameter individual, and apply the parameters contained in the optimal second parameter individual to the second initial rule to obtain the second prediction rule.

[0073] In a possible implementation, based on the predicted solar power supply and the predicted regional power consumption, a power supply adequacy evaluation is performed on the solar power supply system on rainy days to obtain a power supply adequacy evaluation result, including:

[0074] It is determined whether the difference obtained by subtracting the predicted regional power consumption from the predicted solar power supply is greater than a preset power supply adequacy threshold. If so, the power supply adequacy assessment result is determined to be sufficient power supply; otherwise, the power supply adequacy assessment result is determined to be insufficient power supply.

[0075] In a possible implementation manner, after evaluating the power supply adequacy of the solar power supply system on a rainy day, the method further includes:

[0076] Determine whether the power supply adequacy assessment result is of insufficient type. If so, generate an insufficient power supply warning and transmit the insufficient power supply warning to the device or system specified by the staff. Otherwise, repeat the solar power supply system power supply adequacy assessment method.

[0077] The present invention provides a method for evaluating the power supply adequacy of a solar power supply system on rainy days, which can monitor the operating status of the solar power supply system in real time and predict the power load of the solar power supply area, thereby accurately evaluating the power supply adequacy under rainy weather conditions, and facilitating users to adjust their power demand according to the power supply adequacy result and achieve energy optimization configuration; the method has the advantages of simple operation, strong practicality, wide application range, etc. BRIEF DESCRIPTION OF THE DRAWINGS

[0078] The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate embodiments consistent with the invention and, together with the description, serve to explain the principles of the invention.

[0079] Figure 1 A flowchart of a method for evaluating power adequacy of a solar power supply system on rainy days is provided for an embodiment of the present invention.

[0080] Figure 2 A flowchart for obtaining a first prediction rule is provided for an embodiment of the present invention.

[0081] Figure 3 A flowchart for obtaining a second prediction rule is provided for an embodiment of the present invention.

[0082] The above drawings have shown clear embodiments of the present invention, which will be described in more detail below. These drawings and text descriptions are not intended to limit the scope of the present invention in any way, but to illustrate the concept of the present invention for those skilled in the art by referring to specific embodiments. DETAILED DESCRIPTION

[0083] Exemplary embodiments will be described in detail herein, examples of which are shown in the accompanying drawings. When the following description refers to the drawings, the same numbers in different drawings represent the same or similar elements unless otherwise indicated. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with the present invention. Instead, they are merely examples of devices and methods consistent with some aspects of the present invention as detailed in the appended claims.

[0084] The embodiments of the present invention are described in detail below with reference to the accompanying drawings.

[0085] like Figure 1 As shown, an embodiment of the present invention provides a method for evaluating power supply adequacy of a solar power supply system for rainy days, comprising:

[0086] S101, collecting solar power supply influencing factors and daily power supply under rainy days, and using GISA-BP algorithm to learn the solar power supply influencing factors and daily power supply to determine a first prediction rule;

[0087] In one possible implementation, the solar power supply influencing factors and daily power supply under rainy day conditions are collected, including: collecting the average temperature, daily temperature difference, average relative humidity, average wind speed, sunshine hours, visibility, global horizontal radiation, global oblique radiation, diffuse horizontal radiation and diffuse oblique radiation under rainy day conditions to obtain the solar power supply influencing factors; collecting the daily power supply under rainy day conditions.

[0088] It is worth noting that, in addition to the above-mentioned solar power supply influencing factors, other influencing factors can also be used to predict solar power generation to make the solar power generation prediction more accurate.

[0089] In an embodiment of the present invention, a prediction rule is constructed using the BP neural network algorithm, and after updating the parameters of the prediction rule using the GISA algorithm (Global Information Sharing Algorithm) provided by the embodiment of the present invention, a first prediction rule capable of predicting the solar power supply is obtained.

[0090] Optionally, in addition to predicting the solar power supply, the solar power generation can also be predicted, and the power supply can be converted through the solar power generation to achieve the prediction of the solar power supply.

[0091] S102. Collect the regional power consumption characteristics and daily power consumption corresponding to the solar power supply area, and use the GISA-BP algorithm to learn the regional power consumption characteristics and daily power consumption to determine a second prediction rule.

[0092] In a possible implementation manner, collecting the regional power consumption characteristics and daily power consumption corresponding to the solar power supply area includes: collecting the seasonal data, weather data, temperature data, and date type data of the power supply area to obtain the regional power consumption characteristics corresponding to the solar power supply area; collecting the daily power consumption corresponding to the regional power consumption characteristics; where the seasonal data includes spring, summer, autumn, or winter, the weather data includes rain, sunny, cloudy, or snowy, and the date type data includes weekdays or non-weekdays.

[0093] It should be noted that in addition to the above regional power consumption characteristics, other power consumption characteristics can also be used to make the prediction of regional power consumption more accurate.

[0094] To enable those skilled in the art of the present invention to better understand the technical solutions described in the embodiments of the present invention, a brief description of using the GISA-BP algorithm to learn the regional power consumption characteristics and daily power consumption is as follows: It can be assumed that the regional power consumption characteristics and daily power consumption for B days are collected. A vector can be constructed using the daily power consumption for C (C < B) days and the regional power consumption characteristics on the (C + 1)-th day to obtain the input data, and the daily power consumption on the (C + 1)-th day is used as the expected input, and the GISA-BP algorithm is used for learning, so that a second prediction rule can be obtained.

[0095] S103. Collect the real-time solar power supply impact factors on rainy and cloudy days, and use the first prediction rule to analyze the real-time solar power supply factors to determine the predicted solar power supply.

[0096] Collect the real-time solar power supply impact factors on rainy and cloudy days, and construct the real-time solar power supply impact factors as the input of the first prediction rule, so that the solar power supply can be predicted.

[0097] S104, collecting real-time regional power consumption characteristics corresponding to the solar power supply area, and using the second prediction rule to analyze the real-time regional power consumption characteristics to determine the predicted regional power consumption;

[0098] By collecting the real-time regional electricity consumption characteristics corresponding to the solar power supply area, the input data of the second prediction rule can be constructed in a training manner, and the input data of the second prediction rule is input into the second prediction rule, so that the regional electricity consumption can be predicted.

[0099] S105: Based on the predicted solar power supply and the predicted regional power consumption, a power supply adequacy assessment is performed on the solar power supply system on rainy days to obtain a power supply adequacy assessment result.

[0100] When the predicted solar power supply is approximately equal to the predicted regional general electricity consumption, it can be considered that the power supply is basically sufficient. However, electricity consumption will fluctuate. Therefore, a preset power supply surplus threshold can be preset. When the difference between the predicted solar power supply and the predicted regional general electricity consumption is greater than the preset power supply surplus threshold, it can be considered that the power supply demand is met.

[0101] Optionally, since a general solar power supply system is also provided with a storage battery, the power supply of the storage battery may also be considered to make the power supply adequacy evaluation more accurate.

[0102] It is worth noting that in order to ensure that the power supply adequacy assessment method for a solar power supply system on rainy days provided in an embodiment of the present invention can achieve accurate power supply adequacy assessment, a large number of samples are required for data learning, and after the method is deployed, historical data can be continuously learned to correct the assessment effect, thereby making the assessment more and more accurate.

[0103] The present invention provides a method for evaluating the power supply adequacy of a solar power supply system on rainy days, which can monitor the operating status of the solar power supply system in real time and predict the power load of the solar power supply area, thereby accurately evaluating the power supply adequacy under rainy weather conditions, and facilitating users to adjust their power demand according to the power supply adequacy result and achieve energy optimization configuration; the method has the advantages of simple operation, strong practicality, wide application range, etc.

[0104] In the existing deep learning process, in order to achieve the learning of data relationships, intelligent optimization algorithms such as particle swarm optimization algorithm and genetic algorithm are often used to optimize the parameters of the neural network algorithm, so as to achieve data relationship learning. However, the intelligent optimization algorithms in the prior art often have the problems of poor training accuracy, slow training speed and easy to fall into local optimality, which usually leads to poor data relationship learning ability and ultimately makes it difficult to achieve accurate power supply adequacy assessment. Therefore, the embodiment of the present invention provides a GISA-BP algorithm to solve the technical problems existing in the prior art, so as to achieve accurate power supply adequacy assessment.

[0105] like Figure 2 As shown, the GISA-BP (global information sharing-back propagation) algorithm is used to learn the solar power supply influencing factors and daily power supply to determine the first prediction rule, including:

[0106] S201, using the BP algorithm to create a first initial rule, and initialize the parameters of the first initial rule to obtain a first parameter individual, and repeatedly obtain multiple different first parameter individuals;

[0107] In addition to using the BP algorithm to construct the first initial rule, other neural networks can also be used to construct the first initial rule. The parameters of the first initial rule can be initialized using a random initialization method or a chaotic mapping initialization method. The embodiment of the present invention preferably uses a chaotic mapping initialization method to initialize, so that the initial solution is more evenly distributed in the solution space, thereby improving the training speed of the algorithm.

[0108] The parameters of the first initial rule may be connection weights and / or thresholds, so that the first initial rule can implement data analysis and prediction.

[0109] S202, forming a vector of the solar power supply influencing factors and using it as input data of the first initial rule, using the daily power supply corresponding to the solar power supply influencing factors as the expected label, and obtaining the loss function value corresponding to each first parameter individual;

[0110] Obtaining the loss function value corresponding to each first parameter individual may include: using a root mean square error loss function or a cross entropy error loss function to obtain the loss function value.

[0111] S203, determining the optimal first parameter individual according to the loss function value corresponding to each first parameter individual;

[0112] That is, the first parameter individual with the smallest loss function value is determined as the optimal first parameter individual.

[0113] S204, based on the optimal first parameter individual, each first parameter individual is updated by adopting an adaptive vibration-guided search strategy to obtain a first parameter individual after vibration-guided search;

[0114] S205, for the first parameter individuals after the vibration-guided search, adopt an adaptive local area search strategy to update each first parameter individual to obtain the first parameter individual after the local area search;

[0115] S206, for the first parameter individuals after the local area search, each first parameter individual is updated using an adaptive population collaborative search strategy to obtain the first parameter individuals after the population collaborative search;

[0116] S207, repeatedly executing the vibration-guided search, local area search and population collaborative search until the number of training times reaches the maximum number of training times, re-obtaining the optimal first parameter individual, and applying the parameters contained in the optimal first parameter individual to the first initial rule to obtain the first prediction rule.

[0117] Optionally, in addition to terminating the training when the number of training times reaches the maximum number of training times, other conditions may be used to terminate the training. For example, when the minimum loss function value is less than a preset threshold, it may be determined that the training can be terminated.

[0118] In a possible implementation, based on the optimal first parameter individual, each first parameter individual is updated using an adaptive vibration-guided search strategy to obtain a first parameter individual after vibration-guided search, including:

[0119] Based on the current number of training times, the adaptive attenuation factor is determined as:

[0120]

[0121] Among them, z t represents the adaptive attenuation factor during the t-th training process, z max Indicates the maximum value corresponding to the adaptive attenuation factor, z min represents the minimum value corresponding to the adaptive attenuation factor, e represents the natural constant, and T represents the maximum number of training times;

[0122] According to the adaptive attenuation factor, the first adaptive vibration induction coefficient and the second adaptive vibration induction coefficient are obtained as follows:

[0123] α 1 =2z t r 1 -z t

[0124] α 2 =kzt r 2 +1

[0125] Among them, α 1 represents the first adaptive vibration guidance coefficient, r 1 Represents the first random number between (0, 1), α 2 represents the second adaptive vibration guidance coefficient, k represents the natural constant, r 2 represents a second random number between (0, 1);

[0126] Generate a third random number r between (0, 1) 3 , and determine the third random number r 3 Is it less than the first decision probability (which can be set to 0.5), if so, based on the optimal first parameter individual, and according to the first adaptive vibration guidance coefficient and the second adaptive vibration guidance, the first parameter individual is searched for by the first vibration guidance to obtain the first parameter individual after the vibration guidance search, otherwise based on the optimal first parameter individual, and according to the first adaptive vibration guidance coefficient and the second adaptive vibration guidance, the first parameter individual is searched for by the second vibration guidance;

[0127] Based on the optimal first parameter individual, and according to the first adaptive vibration guidance coefficient and the second adaptive vibration guidance, the first parameter individual is searched for the first vibration guidance as follows:

[0128]

[0129] in, represents the i-th first parameter individual in the t-th training process, i=1,2,…,I, I represents the total number of first parameter individuals, represents the optimal first parameter individual, Represents the first parameter individual after vibration-guided search

[0130] Based on the optimal first parameter individual, and according to the first adaptive vibration guidance coefficient and the second adaptive vibration guidance, the second vibration guidance search is performed on the first parameter individual as follows:

[0131]

[0132] Among them, π represents pi, cos represents cosine function, and sin represents sine function.

[0133] The first vibration-guided search and the second vibration-guided search provided in the embodiments of the present invention can enable all first parameter individuals to perform spiral searches in the solution space in different ways and vibrate during the spiral search process, thereby forming a mesh search path, which can not only effectively improve the search efficiency, but also improve the algorithm's global search capability, thereby effectively avoiding falling into local optimality.

[0134] In a possible implementation, for the first parameter individuals after the vibration-guided search, an adaptive local area search strategy is used to update each first parameter individual to obtain the first parameter individual after the local area search, including:

[0135] Based on the current number of training times, the adaptive local development factor is determined as:

[0136] p=(1-t / T) t / T

[0137] Where p represents the adaptive local development factor;

[0138] For the first parameter individual after the vibration-guided search, generate a fourth random number r between (0, 1) 4 , and determine the fourth random number r 4 Is it less than the second decision probability (which can be set to 0.5), if so, perform a first local area search on the first parameter individual to obtain the first parameter individual after the local area search, otherwise perform a second local area search on the first parameter individual to obtain the first parameter individual after the local area search;

[0139] The first local area search for the first parameter individual is:

[0140]

[0141] in, represents the first parameter individual after the nth vibration-guided search, n = 1, 2, ..., I, r 5 represents the fifth random number between (0, 1), and F represents the random quantity between [-1, 1];

[0142] The second local area search for the first parameter individual is:

[0143]

[0144] in, Represents the first parameter individual after the local area search

[0145] The first local area search and the second local area search provided by the embodiment of the present invention can enable the first parameter individual to perform local search in irregular routes and in different ways, thereby effectively improving the search accuracy of the algorithm.

[0146] In a possible implementation, for the first parameter individuals after the local area search, an adaptive population collaborative search strategy is used to update each first parameter individual to obtain the first parameter individuals after the population collaborative search, including:

[0147] Based on the current number of training times, the adaptive cooperation factor is determined as:

[0148]

[0149] in, represents the adaptive cooperation factor corresponding to the first parameter individual after the mth local area search in the tth training process, β 0 represents the first preset constant, represents the intermediate coefficient corresponding to the first parameter individual after the mth local area search, m = 1, 2, ..., I, ξ max represents the maximum value corresponding to the intermediate coefficient, ξ max represents the minimum value corresponding to the intermediate coefficient, It represents the fitness value corresponding to the first parameter individual after the mth local area search. represents the minimum fitness value in the first parameter individual after the local area search, represents the average fitness value of the first parameter individual after the local area search; the fitness value is obtained by adding the loss function value and the second preset constant and taking the reciprocal;

[0150] For the first parameter individual after the mth local area search, according to the adaptive cooperation factor, the cooperation influence of all other first parameter individuals on the first parameter individual is determined as follows:

[0151]

[0152] in, represents the collaborative influence, r 6 represents the sixth random number between (0, 1), Represents the first parameter individual after the kth local area search The first parameter individual after the mth local area search Regarding the influence of the d-th dimension parameter, d=1, 2, ..., D, D represents the total dimension of the parameters of the first parameter individual, ε represents the second preset constant, and ε=0.0001; Represents the first parameter individual The corresponding weights, Represents the first parameter individual The corresponding weights, and The method of obtaining is the same; Represents the first parameter individual The d-th dimension parameter of Represents the first parameter individual The d-th dimension parameter of Represents the first parameter individual The corresponding position quality, Represents the first parameter individual The corresponding position quality, f m represents the fitness value corresponding to the first parameter individual after the mth local area search, f w represents the minimum fitness value of the first parameter individual after the local area search, f best Represents the maximum fitness value of the first parameter individual after the local area search;

[0153] Based on collaborative influence And weight Determine the first parameter individual The collaborative influence term corresponding to the d-th dimension parameter is:

[0154] According to the first parameter The collaborative influence term corresponding to the d-th dimension parameter Determine the first parameter individual The incremental term corresponding to the d-th dimension parameter is:

[0155]

[0156] in, Indicates the first parameter individual in the tth training process The incremental term corresponding to the d-th dimension parameter of Indicates the first parameter individual in the t+1th training process The incremental term corresponding to the d-th dimension parameter, r 7 Represents the seventh random number between (0,1), A 1 Represents the first constant between (0,1), r 8 Represents the eighth random number between (-1,1), represents the d-th dimension parameter of the optimal first parameter individual, A 2 Represents the second constant between (0,1), r 9 Represents the ninth random number between (-1,1);

[0157] Here Indicates that except for the first parameter individual The d-th dimension parameter of a random first parameter individual.

[0158] According to the first parameter The incremental term corresponding to the d-th dimension parameter updates the first parameter individual as follows:

[0159]

[0160] in, Represents the first parameter individual after the population collaborative search The d-th dimension parameter of .

[0161] The adaptive local area search strategy provided by the embodiment of the present invention can not only integrate the information of all first parameter individuals, but also learn the parameter information of the optimal first parameter individual and a random first parameter individual, and adopt a memory strategy, which is more helpful to use the entire population information to explore the unknown area of ​​the entire solution space and improve the possibility of finding the global optimal solution.

[0162] Optionally, after the first parameter individual is updated, out-of-bounds processing may be performed on it, thereby ensuring that the parameter is always valid.

[0163] like Figure 3 As shown, the GISA-BP algorithm is used to learn the regional power consumption characteristics and daily power consumption to determine the second prediction rule, including:

[0164] S301, using the BP algorithm to create a second initial rule, and initialize the parameters of the second initial rule to obtain a second parameter individual, and repeatedly obtain multiple different second parameter individuals;

[0165] S302, the regional electricity consumption features are formed into a vector and used as input data of the second initial rule, the daily electricity consumption corresponding to the regional electricity consumption features is used as the expected label, and the loss function value corresponding to each second parameter individual is obtained;

[0166] S303, determining the optimal second parameter individual according to the loss function value corresponding to each second parameter individual;

[0167] S304, based on the optimal second parameter individual, each second parameter individual is updated by adopting an adaptive vibration-guided search strategy to obtain a second parameter individual after vibration-guided search;

[0168] S305, for the second parameter individuals after the vibration-guided search, an adaptive local area search strategy is used to update each second parameter individual to obtain the second parameter individuals after the local area search;

[0169] S306, for the second parameter individuals after the local area search, an adaptive population collaborative search strategy is used to update each second parameter individual to obtain the second parameter individuals after the population collaborative search;

[0170] S307, repeatedly perform the vibration-guided search, local area search and population collaborative search until the number of training times reaches the maximum number of training times, re-obtain the optimal second parameter individual, and apply the parameters contained in the optimal second parameter individual to the second initial rule to obtain the second prediction rule.

[0171] The specific method of determining the second prediction rule is similar to the principle and beneficial effects of determining the first prediction rule, and will not be repeated here.

[0172] In a possible implementation, based on the predicted solar power supply and the predicted regional power consumption, a power supply adequacy evaluation is performed on the solar power supply system on rainy days to obtain a power supply adequacy evaluation result, including:

[0173] It is determined whether the difference obtained by subtracting the predicted regional power consumption from the predicted solar power supply is greater than a preset power supply adequacy threshold. If so, the power supply adequacy assessment result is determined to be sufficient power supply; otherwise, the power supply adequacy assessment result is determined to be insufficient power supply.

[0174] In a possible implementation manner, after evaluating the power supply adequacy of the solar power supply system on a rainy day, the method further includes:

[0175] Determine whether the power supply adequacy assessment result is of insufficient type. If so, generate an insufficient power supply warning and transmit the insufficient power supply warning to the device or system specified by the staff. Otherwise, repeat the solar power supply system power supply adequacy assessment method.

[0176] It will be appreciated by those skilled in the art that embodiments of the present invention may be provided as methods, systems or computer program products. Therefore, the present invention may take the form of a complete hardware embodiment, a complete software embodiment or an embodiment combining software and hardware. Moreover, the present invention may take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program codes. The schemes in the embodiments of the present invention may be implemented in various computer languages, for example, object-oriented programming language Java and literal scripting language JavaScript, etc.

[0177] The present invention is described with reference to flowcharts and / or block diagrams of methods, devices (systems) and computer program products according to embodiments of the present invention. It should be understood that each process and / or block in the flowchart and / or block diagram and the combination of processes and / or blocks in the flowchart and / or block diagram can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the process in the flowchart. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0178] These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.

[0179] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process. Figure 1 A process or multiple processes and / or boxes Figure 1 The steps for the functions specified in one or more boxes.

[0180] Although the preferred embodiments of the present invention have been described, those skilled in the art may make other changes and modifications to these embodiments once they have learned the basic creative concept. Therefore, the appended claims are intended to be interpreted as including the preferred embodiments and all changes and modifications that fall within the scope of the present invention.

[0181] Obviously, those skilled in the art can make various changes and modifications to the present invention without departing from the spirit and scope of the present invention. Thus, if these modifications and variations of the present invention fall within the scope of the claims of the present invention and their equivalents, the present invention is also intended to include these modifications and variations.

Claims

1. A method for evaluating the power supply adequacy of a solar power supply system on rainy days, characterized in that: include: Collecting the solar power supply influencing factors and daily power supply under rainy days, and using the GISA-BP algorithm to learn the solar power supply influencing factors and daily power supply to determine a first prediction rule; Collecting regional power consumption characteristics and daily power consumption corresponding to the solar power supply area, and using the GISA-BP algorithm to learn the regional power consumption characteristics and daily power consumption to determine a second prediction rule; Collecting real-time solar power supply influencing factors on rainy days, and analyzing the real-time solar power supply factors using the first prediction rule to determine the predicted solar power supply amount; Collecting real-time regional electricity consumption characteristics corresponding to the solar power supply area, and using the second prediction rule to analyze the real-time regional electricity consumption characteristics to determine the predicted regional electricity consumption; According to the predicted solar power supply and the predicted regional power consumption, a power supply adequacy assessment is performed on the solar power supply system on rainy days to obtain a power supply adequacy assessment result.

2. The method for evaluating the power supply adequacy of a solar power supply system for rainy days according to claim 1 is characterized in that: Collect the solar power supply influencing factors and daily power supply under rainy and cloudy conditions, including: The average temperature, daily temperature range, average relative humidity, average wind speed, sunshine hours, visibility, global horizontal radiation, global oblique radiation, diffuse horizontal radiation and diffuse oblique radiation under rainy and cloudy conditions were collected to obtain the solar power supply impact factor; Collect daily power supply under rainy and cloudy conditions.

3. The method for evaluating the power supply adequacy of a solar power supply system for rainy days according to claim 1, characterized in that: The GISA-BP algorithm is used to learn the solar power supply influencing factors and the daily power supply to determine the first prediction rule, including: A BP algorithm is used to create a first initial rule, and the parameters of the first initial rule are initialized to obtain a first parameter individual, and a plurality of different first parameter individuals are repeatedly obtained; The solar power supply influencing factors are formed into vectors and used as input data of the first initial rule. The daily power supply corresponding to the solar power supply influencing factors is used as the expected label, and the loss function value corresponding to each first parameter individual is obtained; Determine the optimal first parameter individual according to the loss function value corresponding to each first parameter individual; Based on the optimal first parameter individual, each first parameter individual is updated by adopting an adaptive vibration-guided search strategy to obtain a first parameter individual after vibration-guided search; For the first parameter individuals after the vibration-guided search, an adaptive local area search strategy is adopted to update each first parameter individual to obtain the first parameter individual after the local area search; For the first parameter individuals after the local area search, an adaptive population collaborative search strategy is used to update each first parameter individual to obtain the first parameter individuals after the population collaborative search; Repeat the vibration-guided search, local area search, and population collaborative search until the number of training times reaches the maximum number of training times, re-obtain the optimal first parameter individual, and apply the parameters contained in the optimal first parameter individual to the first initial rule to obtain the first prediction rule.

4. The method for evaluating the power supply adequacy of a solar power supply system for rainy days according to claim 3 is characterized in that: Based on the optimal first parameter individual, each first parameter individual is updated by adopting an adaptive vibration guided search strategy to obtain a first parameter individual after vibration guided search, including: Based on the current number of training times, the adaptive attenuation factor is determined as: Among them, z t represents the adaptive attenuation factor during the t-th training process, z max Indicates the maximum value corresponding to the adaptive attenuation factor, z min represents the minimum value corresponding to the adaptive attenuation factor, e represents the natural constant, and T represents the maximum number of training times; According to the adaptive attenuation factor, the first adaptive vibration induction coefficient and the second adaptive vibration induction coefficient are obtained as follows: α1=2z t r1-z t α2=kz t r2+1 Wherein, α1 represents the first adaptive vibration induction coefficient, r1 represents the first random number between (0, 1), α2 represents the second adaptive vibration induction coefficient, k represents the natural constant number, and r2 represents the second random number between (0, 1); Generate a third random number r3 between (0, 1), and determine whether the third random number r3 is less than the first decision probability. If so, based on the optimal first parameter individual and according to the first adaptive vibration guidance coefficient and the second adaptive vibration guidance, perform a first vibration guidance search on the first parameter individual to obtain the first parameter individual after the vibration guidance search. Otherwise, based on the optimal first parameter individual and according to the first adaptive vibration guidance coefficient and the second adaptive vibration guidance, perform a second vibration guidance search on the first parameter individual. Based on the optimal first parameter individual, and according to the first adaptive vibration guidance coefficient and the second adaptive vibration guidance, the first parameter individual is searched for the first vibration guidance as follows: in, represents the i-th first parameter individual in the t-th training process, i=1,2,…,I, I represents the total number of first parameter individuals, represents the optimal first parameter individual, Represents the first parameter individual after vibration-guided search Based on the optimal first parameter individual, and according to the first adaptive vibration guidance coefficient and the second adaptive vibration guidance, the second vibration guidance search is performed on the first parameter individual as follows: Among them, π represents pi, cos represents cosine function, and sin represents sine function.

5. The method for evaluating the power supply adequacy of a solar power supply system for rainy days according to claim 4 is characterized in that: For the first parameter individuals after the vibration-guided search, an adaptive local area search strategy is used to update each first parameter individual to obtain the first parameter individual after the local area search, including: Based on the current number of training times, the adaptive local development factor is determined as: p=(1-t / T) t / T Where p represents the adaptive local development factor; For the first parameter individual after the vibration-guided search, a fourth random number r4 between (0, 1) is generated, and it is determined whether the fourth random number r4 is less than the second decision probability. If so, a first local area search is performed on the first parameter individual to obtain the first parameter individual after the local area search. Otherwise, a second local area search is performed on the first parameter individual to obtain the first parameter individual after the local area search. The first local area search for the first parameter individual is: in, represents the first parameter individual after the nth vibration guided search, n=1,2,…,I, r5 represents the fifth random number between (0,1), and F represents the random quantity between [-1,1]; The second local area search for the first parameter individual is: in, Represents the first parameter individual after the local area search 6. The method for evaluating the power supply adequacy of a solar power supply system for rainy days according to claim 5, characterized in that: For the first parameter individuals after the local area search, an adaptive population collaborative search strategy is used to update each first parameter individual to obtain the first parameter individual after the population collaborative search, including: Based on the current number of training times, the adaptive cooperation factor is determined as: in, represents the adaptive cooperation factor corresponding to the first parameter individual after the mth local area search in the tth training process, β0 represents the first preset constant, represents the intermediate coefficient corresponding to the first parameter individual after the mth local area search, m = 1, 2, ..., I, ξ max represents the maximum value corresponding to the intermediate coefficient, ξ max represents the minimum value corresponding to the intermediate coefficient, It represents the fitness value corresponding to the first parameter individual after the mth local area search. represents the minimum fitness value in the first parameter individual after the local area search, represents the average fitness value of the first parameter individual after the local area search; the fitness value is obtained by adding the loss function value and the second preset constant and taking the reciprocal; For the first parameter individual after the mth local area search, according to the adaptive cooperation factor, the cooperation influence of all other first parameter individuals on the first parameter individual is determined as follows: in, represents the collaborative influence, r6 represents the sixth random number between (0, 1), Represents the first parameter individual after the kth local area search The first parameter individual after the mth local area search Regarding the influence of the d-th dimension parameter, d=1, 2, ..., D, D represents the total dimension of the parameters of the first parameter individual, ε represents the second preset constant, and ε=0.0001; Represents the first parameter individual The corresponding weight, Represents the first parameter individual The corresponding weight, and The method of obtaining is the same; Represents the first parameter individual The d-th dimension parameter of Represents the first parameter individual The d-th dimension parameter of Represents the first parameter individual The corresponding position quality, Represents the first parameter individual The corresponding position quality, f m represents the fitness value corresponding to the first parameter individual after the mth local area search, f w represents the minimum fitness value of the first parameter individual after the local area search, f best Represents the maximum fitness value of the first parameter individual after the local area search; Based on collaborative influence And weight Determine the first parameter individual The collaborative influence term corresponding to the d-th dimension parameter is: According to the first parameter The collaborative influence term corresponding to the d-th dimension parameter Determine the first parameter individual The incremental term corresponding to the d-th dimension parameter is: in, Indicates the first parameter individual in the tth training process The incremental term corresponding to the d-th dimension parameter of Indicates the first parameter individual in the t+1th training process The increment item corresponding to the d-th dimension parameter, r7 represents the seventh random number between (0,1), A1 represents the first constant between (0,1), r8 represents the eighth random number between (-1,1), represents the d-th dimension parameter of the optimal first parameter individual, A2 represents the second constant between (0,1), and r9 represents the ninth random number between (-1,1); According to the first parameter The incremental term corresponding to the d-th dimension parameter updates the first parameter individual as follows: in, Represents the first parameter individual after the population collaborative search The d-th dimension parameter of .

7. The method for evaluating the power supply adequacy of a solar power supply system for rainy days according to claim 6, characterized in that: Collect regional electricity consumption characteristics and daily electricity consumption corresponding to the solar power supply area, including: Collect seasonal data, weather data, temperature data and date type data of the power supply area to obtain regional electricity consumption characteristics corresponding to the solar power supply area; Collect the daily power consumption corresponding to the power consumption characteristics of the area; The season data includes spring, summer, autumn or winter, the weather data includes rainy, sunny, cloudy or snowy, and the date type data includes working days or non-working days.

8. The method for evaluating the power supply adequacy of a solar power supply system for rainy days according to claim 7, characterized in that: The GISA-BP algorithm is used to learn the regional power consumption characteristics and daily power consumption to determine the second prediction rule, including: A BP algorithm is used to create a second initial rule, and the parameters of the second initial rule are initialized to obtain a second parameter individual, and a plurality of different second parameter individuals are repeatedly obtained; The regional electricity consumption features are formed into vectors and used as input data for the second initial rule. The daily electricity consumption corresponding to the regional electricity consumption features is used as the expected label to obtain the loss function value corresponding to each second parameter individual; Determine the optimal second parameter individual according to the loss function value corresponding to each second parameter individual; Based on the optimal second parameter individual, each second parameter individual is updated by adopting an adaptive vibration guided search strategy to obtain a second parameter individual after vibration guided search; For the second parameter individuals after the vibration-guided search, an adaptive local area search strategy is adopted to update each second parameter individual to obtain the second parameter individuals after the local area search; For the second parameter individuals after the local area search, an adaptive population collaborative search strategy is used to update each second parameter individual to obtain the second parameter individuals after the population collaborative search; Repeat the vibration-guided search, local area search and population collaborative search until the number of training times reaches the maximum number of training times, re-obtain the optimal second parameter individual, and apply the parameters contained in the optimal second parameter individual to the second initial rule to obtain the second prediction rule.

9. The method for evaluating the power supply adequacy of a solar power supply system for rainy days according to claim 8, characterized in that: According to the predicted solar power supply and the predicted regional power consumption, the power supply adequacy of the solar power supply system on rainy days is evaluated to obtain a power supply adequacy evaluation result, including: It is determined whether the difference obtained by subtracting the predicted regional power consumption from the predicted solar power supply is greater than a preset power supply adequacy threshold. If so, the power supply adequacy assessment result is determined to be sufficient power supply; otherwise, the power supply adequacy assessment result is determined to be insufficient power supply.

10. The method for evaluating the power supply adequacy of a solar power supply system for rainy days according to any one of claims 1 to 9, characterized in that: After evaluating the power supply adequacy of the solar power supply system on rainy days, it also includes: Determine whether the power supply adequacy assessment result is of insufficient type. If so, generate an insufficient power supply warning and transmit the insufficient power supply warning to the device or system specified by the staff. Otherwise, repeat the solar power supply system power supply adequacy assessment method.

Images