A design optimization method and system for arc-free circuit breakers in multiple photovoltaic access areas

Abstract

The present invention provides a method and system for optimizing the design of arc-free circuit breakers in a multi-photovoltaic access lower area, belonging to the technical field of arc-free circuit breakers, and the method includes: obtaining the capacitance value, resistance value, attenuation time constant and circuit-breaking time of the circuit breaker circuit, performing surface fitting operations respectively, obtaining a first reference domain and a second reference domain, and obtaining corresponding reference data based on the first reference domain and the second reference domain, and performing neural network training based on the reference data to obtain a neural network model, and optimizing the neural network model by using a surface fitting method to obtain an optimal neural network model, and finally optimizing the design of arc-free circuit breakers in a multi-photovoltaic access lower area based on the optimal neural network model. The present invention uses surface fitting and neural network methods to optimize the design of arc-free circuit breakers from multiple dimensions.

Description

Technical Field

[0001] The present invention relates to the technical field of circuit breaker design, and in particular to a method and system for optimizing the design of arc-free circuit breakers for multiple photovoltaic access downstage areas. Background Art

[0002] At present, the design optimization scheme of arc-free circuit breakers in the substation area with multiple photovoltaic access mainly takes into account the characteristics of the photovoltaic system, the safe and stable operation of the power grid, and the arc-free operation of the circuit breaker, as follows:

[0003] 1. Design Background

[0004] Photovoltaic system characteristics: With the large-scale access of distributed photovoltaic systems, problems such as voltage rise, three-phase imbalance, and harmonic interference have arisen, which pose challenges to the safety and stable operation of the power grid.

[0005] Arc-free circuit breakers are required: Traditional circuit breakers may generate arcs during operation, posing a threat to equipment and personnel safety. Therefore, the design of arc-free circuit breakers is of great significance to improving the safety and reliability of power grids.

[0006] 2. Design goals

[0007] Arc suppression: Ensure that the circuit breaker does not generate arcs during operation, reducing equipment and personnel safety risks.

[0008] Improve grid stability: By optimizing the design of circuit breakers, the grid's adaptability to distributed photovoltaic access can be improved to ensure stable operation of the grid.

[0009] 3. Design Method

[0010] 3.1 Choose the right arc-free technology

[0011] Magnetic induction transfer and resistor current limiting technology: Use the induction module to quickly transfer the current to the capacitor and resistor to achieve fast switching without arcing and improve the withstand voltage of the break. At the same time, the resistor current limiting can limit the current to a certain level to achieve current shutoff.

[0012] 3.2 Design a reasonable circuit breaker topology

[0013] Capacitor: used to store energy and maintain voltage stability, slowing down the rate of increase of electric field strength at both ends of the switch contacts.

[0014] Inductor: used to continue the current and suppress the rising speed of short-circuit current.

[0015] Diode: Makes the circuit breaker operate in different states when on and off.

[0016] High-power resistor: discharges the energy stored in the circuit breaker energy storage element.

[0017] Mechanical switch: realizes the on and off of the control circuit, and ensures that the circuit is reliably shut down after the action, realizing the transfer of arc energy.

[0018] 3.3 Optimization control strategy

[0019] High-precision measurement and algorithm optimization: Through high-precision measurement of current, voltage and other parameters, combined with advanced control algorithms, accurate control of the circuit breaker operation process is achieved to ensure arc-free operation.

[0020] Smart Internet of Things technology: Smart Internet of Things technology can be used to achieve remote monitoring and intelligent control of circuit breakers, thereby improving the automation level and reliability of power grid operation.

[0021] 4. Implementation steps

[0022] Demand analysis: clarify the grid characteristics of the area where multiple photovoltaics are connected, the access method and capacity of the distributed photovoltaic system, and other requirements.

[0023] Solution design: Based on the results of demand analysis, design a suitable arc-free circuit breaker solution, including arc-free technology selection, topology design, control strategy optimization, etc.

[0024] Simulation verification: The design scheme is simulated and verified through simulation software to ensure the realization of arc-free operation and stable operation of the power grid.

[0025] Experimental verification: Conduct actual tests in the laboratory or on-site to verify the feasibility and effectiveness of the design solution.

[0026] However, the above-mentioned traditional solutions have been in use with few major changes, which is not conducive to the development of circuit breakers and photovoltaic technologies. Moreover, they are all considered and optimized from a single dimension. Therefore, the present invention attempts to provide a multi-dimensional design optimization solution. Summary of the invention

[0027] The present invention provides a method and system for designing and optimizing arc-free circuit breakers in a multi-photovoltaic access substation, which adopts surface fitting and neural network methods to design and optimize arc-free circuit breakers from multiple dimensions.

[0028] The embodiment of this specification discloses a method for optimizing the design of arc-free circuit breakers in a multi-photovoltaic access substation, including:

[0029] Obtain the capacitance value, resistance value, decay time constant and breaking time of the circuit breaker circuit;

[0030] The capacitance value is used as a row vector, and the decay time constant is used as a column vector to form a first matrix, and then the disconnection time is introduced, and a surface fitting operation is performed to obtain a three-dimensional surface and a capacitance value surface fitting equation, and the three-dimensional surface is recorded as a first reference domain;

[0031] The resistance value is used as a row vector and the decay time constant is used as a column vector to form a second matrix, and then the disconnection time is introduced to perform a surface fitting operation to obtain a three-dimensional surface and a resistance value surface fitting equation, and the three-dimensional surface is recorded as a second reference domain;

[0032] Based on the first reference domain and the second reference domain, with the goal of optimizing the disconnection time, a reference disconnection time, a reference capacitance value, a reference resistance value, and a reference decay time constant are obtained;

[0033] Arrange the reference disconnection time, the reference capacitance value, the reference resistance value, and the reference decay time constant into a first training set and a first test set, and use the reference disconnection time alone as a verification set;

[0034] Preprocessing the first training set and the first test set;

[0035] Based on the preprocessed first training set, model training is performed to obtain a first neural network model;

[0036] Based on the preprocessed first test set, obtaining a first accuracy, a first precision, and a first recall of the first neural network model;

[0037] The first recall rate is used as a row vector, and the first accuracy rate is used as a column vector to form a third matrix, and then the reference disconnection time is imported, and a surface fitting operation is performed to obtain a three-dimensional surface, and the three-dimensional surface is recorded as a third reference domain;

[0038] The first precision is used as a row vector, the first accuracy rate is used as a column vector, a fourth matrix is ​​formed, the reference disconnection time is introduced, a surface fitting operation is performed, a three-dimensional surface is obtained, and the three-dimensional surface is recorded as a fourth reference domain;

[0039] Based on the third reference domain and the fourth reference domain, with the goal of optimizing the reference disconnection time, using the first test set, selecting the optimal accuracy, optimal precision, and optimal recall rate of the model corresponding to the optimal reference disconnection time;

[0040] Based on the optimal accuracy, the optimal precision and the optimal recall rate, the model parameters are optimized by adjusting the hyperparameters, modifying the model structure and using regularization, and based on the validation set, the model parameters corresponding to the optimal reference disconnection time are selected to obtain the optimal first neural network model;

[0041] The design optimization of arc-free circuit breakers in the multi-PV access area is carried out based on the optimal first neural network model.

[0042] In some embodiments of this specification, a rated voltage of a circuit breaker circuit is obtained;

[0043] The rated voltage is used as a row vector and the decay time constant is used as a column vector to form a fifth matrix, and then the disconnection time is introduced to perform a surface fitting operation to obtain a three-dimensional surface and a rated voltage surface fitting equation, and the three-dimensional surface is recorded as a fifth reference domain;

[0044] Based on the first reference domain, the second reference domain and the fifth reference domain, with the goal of optimizing the breaking time, a reference breaking time, a reference capacitance value, a reference resistance value, a reference decay time constant and a reference rated voltage are obtained;

[0045] Arrange the reference disconnection time, the reference capacitance value, the reference resistance value, the reference decay time constant and the reference rated voltage into a second training set and a second test set;

[0046] Preprocessing the second training set and the second test set;

[0047] Based on the preprocessed second training set, model training is performed to obtain a second neural network model;

[0048] Based on the preprocessed second test set, obtaining a second accuracy, a second precision, and a second recall of the second neural network model;

[0049] The second recall rate is used as a row vector, and the second accuracy rate is used as a column vector to form a sixth matrix, and then the reference disconnection time is imported, and a surface fitting operation is performed to obtain a three-dimensional surface, and the three-dimensional surface is recorded as a sixth reference domain;

[0050] The second precision is used as a row vector, and the second accuracy rate is used as a column vector to form a seventh matrix, and then the reference disconnection time is introduced, and a surface fitting operation is performed to obtain a three-dimensional surface, and the three-dimensional surface is recorded as a seventh reference domain;

[0051] Based on the sixth reference domain and the seventh reference domain, with the goal of optimizing the reference disconnection time, using the second test set, selecting the optimal accuracy, optimal precision, and optimal recall rate of the model corresponding to the optimal reference disconnection time;

[0052] Based on the optimal accuracy, the optimal precision and the optimal recall rate, the model parameters are optimized by adjusting the hyperparameters, modifying the model structure and using regularization, and based on the validation set, the model parameters corresponding to the optimal reference disconnection time are selected to obtain the optimal second neural network model;

[0053] The design of arc-free circuit breakers in the multi-PV access area is optimized based on the optimal second neural network model.

[0054] In some embodiments of this specification, a rated current of a circuit breaker circuit is obtained;

[0055] The rated current is used as a row vector and the decay time constant is used as a column vector to form an eighth matrix, and then the disconnection time is introduced to perform a surface fitting operation to obtain a three-dimensional surface and a rated current surface fitting equation, and the three-dimensional surface is recorded as an eighth reference domain;

[0056] Based on the first reference domain, the second reference domain and the eighth reference domain, with the goal of optimizing the breaking time, a reference breaking time, a reference capacitance value, a reference resistance value, a reference decay time constant and a reference rated current are obtained;

[0057] Arrange the reference disconnection time, the reference capacitance value, the reference resistance value, the reference decay time constant and the reference rated current into a third training set and a third test set;

[0058] Preprocessing the third training set and the third test set;

[0059] Based on the preprocessed third training set, model training is performed to obtain a third neural network model;

[0060] Based on the preprocessed third test set, obtaining a third accuracy, a third precision, and a third recall rate of the third neural network model;

[0061] The third recall rate is used as a row vector, and the third accuracy rate is used as a column vector to form a ninth matrix, and then the reference disconnection time is introduced, and a surface fitting operation is performed to obtain a three-dimensional surface, and the three-dimensional surface is recorded as a ninth reference domain;

[0062] The third precision is used as a row vector, and the third accuracy rate is used as a column vector to form a tenth matrix, and then the reference disconnection time is introduced, and a surface fitting operation is performed to obtain a three-dimensional surface, and the three-dimensional surface is recorded as a tenth reference domain;

[0063] Based on the ninth reference domain and the tenth reference domain, with the reference disconnection time being the optimal target, using the third test set, selecting the optimal accuracy, optimal precision and optimal recall rate of the model corresponding to the optimal reference disconnection time;

[0064] Based on the optimal accuracy, the optimal precision and the optimal recall rate, the model parameters are optimized by adjusting the hyperparameters, modifying the model structure and adopting regularization, and based on the validation set, the model parameters corresponding to the optimal reference disconnection time are selected to obtain the optimal third neural network model;

[0065] The design of arc-free circuit breakers in the multi-PV access area is optimized based on the optimal third neural network model.

[0066] In some embodiments of this specification, the power supply voltage fluctuation amplitude of the multi-photovoltaic access lower stage area is obtained;

[0067] The power supply voltage fluctuation amplitude is used as a row vector, and the decay time constant is used as a column vector to form an eleventh matrix, and then the disconnection time is introduced, and a surface fitting operation is performed to obtain a three-dimensional surface and a power supply voltage fluctuation amplitude surface fitting equation, and the three-dimensional surface is recorded as an eleventh reference domain;

[0068] Based on the first reference domain, the second reference domain and the eleventh reference domain, with the goal of optimizing the disconnection time, a reference disconnection time, a reference capacitance value, a reference resistance value, a reference attenuation time constant and a power supply voltage fluctuation amplitude are obtained;

[0069] Arrange the reference disconnection time, the reference capacitance value, the reference resistance value, the reference decay time constant and the power supply voltage fluctuation amplitude into a fourth training set and a fourth test set;

[0070] Preprocessing the fourth training set and the fourth test set;

[0071] Based on the preprocessed fourth training set, model training is performed to obtain a fourth neural network model;

[0072] Based on the preprocessed fourth test set, obtaining a fourth accuracy, a fourth precision, and a fourth recall rate of the fourth neural network model;

[0073] The fourth recall rate is used as a row vector, and the fourth accuracy rate is used as a column vector to form a twelfth matrix, and then the reference disconnection time is introduced, and a surface fitting operation is performed to obtain a three-dimensional surface, and the three-dimensional surface is recorded as a twelfth reference domain;

[0074] The fourth precision is used as a row vector, the fourth accuracy rate is used as a column vector, a thirteenth matrix is ​​formed, and then the reference disconnection time is introduced, a surface fitting operation is performed, and a three-dimensional surface is obtained, and the three-dimensional surface is recorded as a thirteenth reference domain;

[0075] Based on the twelfth reference domain and the thirteenth reference domain, with the goal of optimizing the reference disconnection time, using the fourth test set, selecting the optimal accuracy, optimal precision and optimal recall rate of the model corresponding to the optimal reference disconnection time;

[0076] Based on the optimal accuracy, the optimal precision and the optimal recall rate, the model parameters are optimized by adjusting the hyperparameters, modifying the model structure and adopting regularization, and based on the validation set, the model parameters corresponding to the optimal reference disconnection time are selected to obtain the optimal fourth neural network model;

[0077] The design of arc-free circuit breakers in the multi-PV access area is optimized based on the optimal fourth neural network model.

[0078] In some embodiments of this specification, a current difference between a fault current of a multi-PV access lower section and a rated current of a circuit breaker circuit is obtained;

[0079] The current difference is used as a row vector and the decay time constant is used as a column vector to form a fourteenth matrix, and then the disconnection time is introduced to perform a surface fitting operation to obtain a three-dimensional surface and a current difference surface fitting equation, and the three-dimensional surface is recorded as a fourteenth reference domain;

[0080] Based on the first reference domain, the second reference domain and the fourteenth reference domain, with the goal of optimizing the disconnection time, a reference disconnection time, a reference capacitance value, a reference resistance value, a reference decay time constant and a reference current difference are obtained;

[0081] Arrange the reference disconnection time, the reference capacitance value, the reference resistance value, the reference decay time constant and the reference current difference value into a fifth training set and a fifth test set;

[0082] Preprocessing the fifth training set and the fifth test set;

[0083] Based on the preprocessed fifth training set, model training is performed to obtain a fifth neural network model;

[0084] Based on the preprocessed fifth test set, obtaining a fifth accuracy, a fifth precision, and a fifth recall rate of the fifth neural network model;

[0085] The fifth recall rate is used as a row vector, and the fifth accuracy rate is used as a column vector to form a fifteenth matrix, and then the reference disconnection time is imported, and a surface fitting operation is performed to obtain a three-dimensional surface, and the three-dimensional surface is recorded as a fifteenth reference domain;

[0086] The fifth precision is used as a row vector, the fifth accuracy rate is used as a column vector, a sixteenth matrix is ​​formed, and then the reference disconnection time is introduced, a surface fitting operation is performed, and a three-dimensional surface is obtained, and the three-dimensional surface is recorded as a sixteenth reference domain;

[0087] Based on the fifteenth reference domain and the sixteenth reference domain, with the goal of optimizing the reference disconnection time, using the fifth test set, selecting the optimal accuracy, optimal precision, and optimal recall rate of the model corresponding to the optimal reference disconnection time;

[0088] Based on the optimal accuracy, the optimal precision and the optimal recall rate, the model parameters are optimized by adjusting the hyperparameters, modifying the model structure and adopting regularization, and based on the validation set, the model parameters corresponding to the optimal reference disconnection time are selected to obtain the optimal fifth neural network model;

[0089] The design optimization of arc-free circuit breakers in the multi-PV access area is carried out based on the optimal fifth neural network model.

[0090] In some embodiments of this specification, obtaining an ambient temperature of a circuit breaker circuit;

[0091] The ambient temperature is used as a row vector and the decay time constant is used as a column vector to form a seventeenth matrix, and then the disconnection time is introduced to perform a surface fitting operation to obtain a three-dimensional surface and an ambient temperature surface fitting equation, and the three-dimensional surface is recorded as the seventeenth reference domain;

[0092] Based on the first reference domain, the second reference domain and the seventeenth reference domain, with the goal of optimizing the disconnection time, a reference disconnection time, a reference capacitance value, a reference resistance value, a reference decay time constant and a reference ambient temperature are obtained;

[0093] Arrange the reference disconnection time, the reference capacitance value, the reference resistance value, the reference decay time constant and the reference ambient temperature into a sixth training set and a sixth test set;

[0094] Preprocessing the sixth training set and the sixth test set;

[0095] Based on the preprocessed sixth training set, model training is performed to obtain a sixth neural network model;

[0096] Based on the preprocessed sixth test set, obtaining a sixth accuracy, a sixth precision, and a sixth recall rate of the sixth neural network model;

[0097] The sixth recall rate is used as a row vector, and the sixth accuracy rate is used as a column vector to form an eighteenth matrix, and then the reference disconnection time is introduced, and a surface fitting operation is performed to obtain a three-dimensional surface, and the three-dimensional surface is recorded as the eighteenth reference domain;

[0098] The sixth precision is used as a row vector, the sixth accuracy rate is used as a column vector, a nineteenth matrix is ​​formed, and then the reference disconnection time is introduced, a surface fitting operation is performed, and a three-dimensional surface is obtained, and the three-dimensional surface is recorded as a nineteenth reference domain;

[0099] Based on the eighteenth reference domain and the nineteenth reference domain, with the reference disconnection time being the optimal target, using the sixth test set, selecting the optimal accuracy, optimal precision and optimal recall rate of the model corresponding to the optimal reference disconnection time;

[0100] Based on the optimal accuracy, the optimal precision and the optimal recall rate, the model parameters are optimized by adjusting the hyperparameters, modifying the model structure and adopting regularization, and based on the validation set, the model parameters corresponding to the optimal reference disconnection time are selected to obtain the optimal sixth neural network model;

[0101] The design of arc-free circuit breakers in the multi-PV access area is optimized based on the optimal sixth neural network model.

[0102] In some embodiments of this specification, the harmonic current of the multiple photovoltaic access lower stage area is obtained;

[0103] The harmonic current is used as a row vector and the decay time constant is used as a column vector to form a twentieth matrix, and then the disconnection time is introduced to perform a surface fitting operation to obtain a three-dimensional surface and a harmonic current surface fitting equation, and the three-dimensional surface is recorded as a twentieth reference domain;

[0104] Based on the first reference domain, the second reference domain and the twentieth reference domain, with the goal of optimizing the breaking time, a reference breaking time, a reference capacitance value, a reference resistance value, a reference attenuation time constant and a reference harmonic current are obtained;

[0105] Arrange the reference breaking time, the reference capacitance value, the reference resistance value, the reference decay time constant and the reference harmonic current into a seventh training set and a seventh test set;

[0106] Preprocessing the seventh training set and the seventh test set;

[0107] Based on the preprocessed seventh training set, model training is performed to obtain a seventh neural network model;

[0108] Based on the preprocessed seventh test set, obtaining a seventh accuracy, a seventh precision, and a seventh recall rate of the seventh neural network model;

[0109] The seventh recall rate is used as a row vector, and the seventh accuracy rate is used as a column vector to form a twenty-first matrix, and then the reference disconnection time is imported, and a surface fitting operation is performed to obtain a three-dimensional surface, and the three-dimensional surface is recorded as a twenty-first reference domain;

[0110] The seventh precision is used as a row vector, and the seventh accuracy rate is used as a column vector to form a twenty-second matrix, and then the reference disconnection time is introduced, and a surface fitting operation is performed to obtain a three-dimensional surface, and the three-dimensional surface is recorded as a twenty-second reference domain;

[0111] Based on the twenty-first reference domain and the twenty-second reference domain, with the reference disconnection time being the optimal target, using the seventh test set, selecting the optimal accuracy, optimal precision, and optimal recall rate of the model corresponding to the optimal reference disconnection time;

[0112] Based on the optimal accuracy, the optimal precision and the optimal recall rate, the model parameters are optimized by adjusting the hyperparameters, modifying the model structure and adopting regularization, and based on the validation set, the model parameters corresponding to the optimal reference disconnection time are selected to obtain the optimal seventh neural network model;

[0113] The design optimization of arc-free circuit breakers in the downstation area with multiple photovoltaic accesses is carried out based on the optimal seventh neural network model.

[0114] In some embodiments of this specification, the harmonic voltage of the lower area where multiple photovoltaics are connected is obtained;

[0115] The harmonic voltage is used as a row vector and the decay time constant is used as a column vector to form a twenty-third matrix, and then the disconnection time is introduced to perform a surface fitting operation to obtain a three-dimensional surface and a harmonic voltage surface fitting equation, and the three-dimensional surface is recorded as a twenty-third reference domain;

[0116] Based on the first reference domain, the second reference domain and the twenty-third reference domain, with the goal of optimizing the breaking time, a reference breaking time, a reference capacitance value, a reference resistance value, a reference attenuation time constant and a reference harmonic voltage are obtained;

[0117] Arrange the reference disconnection time, the reference capacitance value, the reference resistance value, the reference decay time constant and the reference harmonic voltage into an eighth training set and an eighth test set;

[0118] Preprocessing the eighth training set and the eighth test set;

[0119] Based on the preprocessed eighth training set, model training is performed to obtain an eighth neural network model;

[0120] Based on the preprocessed eighth test set, obtaining an eighth accuracy, an eighth precision, and an eighth recall rate of the eighth neural network model;

[0121] The eighth recall rate is used as a row vector, and the eighth accuracy rate is used as a column vector to form a twenty-fourth matrix, and then the reference disconnection time is imported, and a surface fitting operation is performed to obtain a three-dimensional surface, and the three-dimensional surface is recorded as a twenty-fourth reference domain;

[0122] The eighth precision is used as a row vector, the eighth accuracy rate is used as a column vector, a twenty-fifth matrix is ​​formed, and then the reference disconnection time is introduced, a surface fitting operation is performed, and a three-dimensional surface is obtained, and the three-dimensional surface is recorded as a twenty-fifth reference domain;

[0123] Based on the twenty-fourth reference domain and the twenty-fifth reference domain, with the reference disconnection time being the optimal target, using the eighth test set, selecting the optimal accuracy, optimal precision, and optimal recall rate of the model corresponding to the optimal reference disconnection time;

[0124] Based on the optimal accuracy, the optimal precision and the optimal recall rate, the model parameters are optimized by adjusting the hyperparameters, modifying the model structure and adopting regularization, and based on the validation set, the model parameters corresponding to the optimal reference disconnection time are selected to obtain the optimal eighth neural network model;

[0125] The design of arc-free circuit breakers in the multi-PV access area is optimized based on the optimal eighth neural network model.

[0126] In some embodiments of this specification, the harmonic frequency of the multiple photovoltaic access lower stage area is obtained;

[0127] The harmonic frequency is used as a row vector and the decay time constant is used as a column vector to form a twenty-sixth matrix, and then the disconnection time is introduced to perform a surface fitting operation to obtain a three-dimensional surface and a harmonic frequency surface fitting equation, and the three-dimensional surface is recorded as the twenty-sixth reference domain;

[0128] Based on the first reference domain, the second reference domain and the twenty-sixth reference domain, with the goal of optimizing the disconnection time, a reference disconnection time, a reference capacitance value, a reference resistance value, a reference attenuation time constant and a reference harmonic frequency are obtained;

[0129] Arrange the reference disconnection time, reference capacitance value, reference resistance value, reference decay time constant and reference harmonic frequency into a ninth training set and a ninth test set;

[0130] Preprocessing the ninth training set and the ninth test set;

[0131] Based on the preprocessed ninth training set, model training is performed to obtain a ninth neural network model;

[0132] Based on the preprocessed ninth test set, obtaining a ninth accuracy, a ninth precision, and a ninth recall rate of the ninth neural network model;

[0133] The ninth recall rate is used as a row vector, and the ninth accuracy rate is used as a column vector to form a twenty-seventh matrix, and then the reference disconnection time is imported, and a surface fitting operation is performed to obtain a three-dimensional surface, and the three-dimensional surface is recorded as a twenty-seventh reference domain;

[0134] The ninth precision is used as a row vector, the ninth accuracy rate is used as a column vector, a twenty-eighth matrix is ​​formed, and then the reference disconnection time is introduced, a surface fitting operation is performed, and a three-dimensional surface is obtained, and the three-dimensional surface is recorded as a twenty-eighth reference domain;

[0135] Based on the twenty-seventh reference domain and the twenty-eighth reference domain, with the reference disconnection time being the optimal target, using the ninth test set, selecting the optimal accuracy, optimal precision, and optimal recall rate of the model corresponding to the optimal reference disconnection time;

[0136] Based on the optimal accuracy, the optimal precision and the optimal recall rate, the model parameters are optimized by adjusting the hyperparameters, modifying the model structure and adopting regularization, and based on the validation set, the model parameters corresponding to the optimal reference disconnection time are selected to obtain the optimal ninth neural network model;

[0137] The design of arc-free circuit breakers in the multi-PV access substation is optimized based on the optimal ninth neural network model.

[0138] The embodiment of this specification also discloses a system for optimizing the design of arc-free circuit breakers in a multi-photovoltaic access lower area, which is used to implement any one of the above-mentioned methods for optimizing the design of arc-free circuit breakers in a multi-photovoltaic access lower area. The system for optimizing the design of arc-free circuit breakers in a multi-photovoltaic access lower area includes:

[0139] An acquisition module, used for acquiring a capacitance value, a resistance value, a decay time constant and a breaking time of a circuit breaker circuit;

[0140] The first fitting module is used to:

[0141] The capacitance value is used as a row vector, and the decay time constant is used as a column vector to form a first matrix, and then the disconnection time is introduced, and a surface fitting operation is performed to obtain a three-dimensional surface and a capacitance value surface fitting equation, and the three-dimensional surface is recorded as a first reference domain;

[0142] The resistance value is used as a row vector and the decay time constant is used as a column vector to form a second matrix, and then the disconnection time is introduced to perform a surface fitting operation to obtain a three-dimensional surface and a resistance value surface fitting equation, and the three-dimensional surface is recorded as a second reference domain;

[0143] A reference module, for obtaining a reference breaking time, a reference capacitance value, a reference resistance value, and a reference attenuation time constant based on the first reference domain and the second reference domain with the goal of optimizing the breaking time;

[0144] Neural network modules for:

[0145] Arrange the reference disconnection time, the reference capacitance value, the reference resistance value, and the reference decay time constant into a first training set and a first test set, and use the reference disconnection time alone as a verification set;

[0146] Preprocessing the first training set and the first test set;

[0147] Based on the preprocessed first training set, model training is performed to obtain a first neural network model;

[0148] Based on the preprocessed first test set, obtaining a first accuracy, a first precision, and a first recall of the first neural network model;

[0149] The second fitting module is used to:

[0150] The first recall rate is used as a row vector, and the first accuracy rate is used as a column vector to form a third matrix, and then the reference disconnection time is imported, and a surface fitting operation is performed to obtain a three-dimensional surface, and the three-dimensional surface is recorded as a third reference domain;

[0151] The first precision is used as a row vector, the first accuracy rate is used as a column vector, a fourth matrix is ​​formed, the reference disconnection time is introduced, a surface fitting operation is performed, a three-dimensional surface is obtained, and the three-dimensional surface is recorded as a fourth reference domain;

[0152] A first optimization module is used to select the optimal accuracy, optimal precision and optimal recall rate of the model corresponding to the optimal reference breaking time based on the third reference domain and the fourth reference domain and taking the optimal reference breaking time as the goal, using the first test set;

[0153] A second optimization module is used to optimize the model parameters by adjusting the hyperparameters, modifying the model structure and adopting regularization based on the optimal accuracy, the optimal precision and the optimal recall rate, and select the model parameters corresponding to the optimal reference disconnection time based on the validation set to obtain the optimal first neural network model;

[0154] The design optimization module is used to optimize the design of arc-free circuit breakers in the multi-PV access downstation area based on the optimal first neural network model.

[0155] The embodiments of this specification can at least achieve the following beneficial effects:

[0156] By obtaining the capacitance value, resistance value, attenuation time constant and breaking time of the circuit breaker circuit, surface fitting operations are performed respectively to obtain the first reference domain and the second reference domain, and based on the first reference domain and the second reference domain, corresponding reference data are obtained, and neural network training is performed based on the reference data to obtain a neural network model, and the surface fitting method is used to optimize the neural network model to obtain the optimal neural network model, and finally the design optimization of arc-free circuit breakers in the multi-photovoltaic access downstage area is performed based on the optimal neural network model. The present invention adopts surface fitting and neural network methods to design and optimize arc-free circuit breakers from multiple dimensions, first establishes a reference domain through historical data or simulation data, then performs neural network training, and uses surface fitting to optimize the neural network model to obtain the optimal neural network model, and finally optimizes the design of arc-free circuit breakers in the multi-photovoltaic access downstage area based on the optimal neural network model, so that the designed arc-free circuit breaker can be more suitable for application in the current multi-photovoltaic access downstage area to obtain an arc-free circuit breaker with the optimal breaking time. BRIEF DESCRIPTION OF THE DRAWINGS

[0157] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings required for use in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other accompanying drawings can be obtained based on these accompanying drawings without paying creative work.

[0158] Figure 1 It is a schematic diagram of the design optimization method of arc-free circuit breaker for multiple photovoltaic access downstage involved in the present invention. DETAILED DESCRIPTION

[0159] In the following, only some exemplary embodiments are briefly described. As those skilled in the art will appreciate, the described embodiments may be modified in various ways without departing from the spirit or scope of the embodiments of the present invention. Therefore, the drawings and descriptions are considered to be exemplary and non-restrictive in nature.

[0160] The disclosure below provides many different embodiments or examples to implement different structures of the embodiments of the present invention. In order to simplify the disclosure of the embodiments of the present invention, the parts and settings of specific examples are described below. Of course, they are only examples, and the purpose is not to limit the embodiments of the present invention. In addition, the embodiments of the present invention can repeat reference numbers and / or reference letters in different examples, and this repetition is for the purpose of simplification and clarity, which itself does not indicate the relationship between the various embodiments and / or settings discussed.

[0161] The embodiments of the present invention are described in detail below with reference to the accompanying drawings.

[0162] like Figure 1 As shown, the embodiment of this specification discloses a method for optimizing the design of arc-free circuit breakers for multiple photovoltaic access downstations, including:

[0163] Obtain the capacitance value, resistance value, decay time constant and breaking time of the circuit breaker circuit;

[0164] The capacitance value is used as a row vector, and the decay time constant is used as a column vector to form a first matrix, and then the disconnection time is introduced, and a surface fitting operation is performed to obtain a three-dimensional surface and a capacitance value surface fitting equation, and the three-dimensional surface is recorded as a first reference domain;

[0165] The resistance value is used as a row vector and the decay time constant is used as a column vector to form a second matrix, and then the disconnection time is introduced to perform a surface fitting operation to obtain a three-dimensional surface and a resistance value surface fitting equation, and the three-dimensional surface is recorded as a second reference domain;

[0166] Based on the first reference domain and the second reference domain, with the goal of optimizing the disconnection time, a reference disconnection time, a reference capacitance value, a reference resistance value, and a reference decay time constant are obtained;

[0167] Arrange the reference disconnection time, the reference capacitance value, the reference resistance value, and the reference decay time constant into a first training set and a first test set, and use the reference disconnection time alone as a verification set;

[0168] Preprocessing the first training set and the first test set;

[0169] Based on the preprocessed first training set, model training is performed to obtain a first neural network model;

[0170] Based on the preprocessed first test set, obtaining a first accuracy, a first precision, and a first recall of the first neural network model;

[0171] The first recall rate is used as a row vector, and the first accuracy rate is used as a column vector to form a third matrix, and then the reference disconnection time is imported, and a surface fitting operation is performed to obtain a three-dimensional surface, and the three-dimensional surface is recorded as a third reference domain;

[0172] The first precision is used as a row vector, the first accuracy rate is used as a column vector, a fourth matrix is ​​formed, the reference disconnection time is introduced, a surface fitting operation is performed, a three-dimensional surface is obtained, and the three-dimensional surface is recorded as a fourth reference domain;

[0173] Based on the third reference domain and the fourth reference domain, with the goal of optimizing the reference disconnection time, using the first test set, selecting the optimal accuracy, optimal precision, and optimal recall rate of the model corresponding to the optimal reference disconnection time;

[0174] Based on the optimal accuracy, the optimal precision and the optimal recall rate, the model parameters are optimized by adjusting the hyperparameters, modifying the model structure and using regularization, and based on the validation set, the model parameters corresponding to the optimal reference disconnection time are selected to obtain the optimal first neural network model;

[0175] The design optimization of arc-free circuit breakers in the multi-PV access area is carried out based on the optimal first neural network model.

[0176] In some embodiments of this specification, a rated voltage of a circuit breaker circuit is obtained;

[0177] The rated voltage is used as a row vector and the decay time constant is used as a column vector to form a fifth matrix, and then the disconnection time is introduced to perform a surface fitting operation to obtain a three-dimensional surface and a rated voltage surface fitting equation, and the three-dimensional surface is recorded as a fifth reference domain;

[0178] Based on the first reference domain, the second reference domain and the fifth reference domain, with the goal of optimizing the breaking time, a reference breaking time, a reference capacitance value, a reference resistance value, a reference decay time constant and a reference rated voltage are obtained;

[0179] Arrange the reference disconnection time, the reference capacitance value, the reference resistance value, the reference decay time constant and the reference rated voltage into a second training set and a second test set;

[0180] Preprocessing the second training set and the second test set;

[0181] Based on the preprocessed second training set, model training is performed to obtain a second neural network model;

[0182] Based on the preprocessed second test set, obtaining a second accuracy, a second precision, and a second recall of the second neural network model;

[0183] The second recall rate is used as a row vector, and the second accuracy rate is used as a column vector to form a sixth matrix, and then the reference disconnection time is imported, and a surface fitting operation is performed to obtain a three-dimensional surface, and the three-dimensional surface is recorded as a sixth reference domain;

[0184] The second precision is used as a row vector, and the second accuracy rate is used as a column vector to form a seventh matrix, and then the reference disconnection time is introduced, and a surface fitting operation is performed to obtain a three-dimensional surface, and the three-dimensional surface is recorded as a seventh reference domain;

[0185] Based on the sixth reference domain and the seventh reference domain, with the goal of optimizing the reference disconnection time, using the second test set, selecting the optimal accuracy, optimal precision, and optimal recall rate of the model corresponding to the optimal reference disconnection time;

[0186] Based on the optimal accuracy, the optimal precision and the optimal recall rate, the model parameters are optimized by adjusting the hyperparameters, modifying the model structure and using regularization, and based on the validation set, the model parameters corresponding to the optimal reference disconnection time are selected to obtain the optimal second neural network model;

[0187] The design of arc-free circuit breakers in the multi-PV access area is optimized based on the optimal second neural network model.

[0188] In some embodiments of this specification, a rated current of a circuit breaker circuit is obtained;

[0189] The rated current is used as a row vector and the decay time constant is used as a column vector to form an eighth matrix, and then the disconnection time is introduced to perform a surface fitting operation to obtain a three-dimensional surface and a rated current surface fitting equation, and the three-dimensional surface is recorded as an eighth reference domain;

[0190] Based on the first reference domain, the second reference domain and the eighth reference domain, with the goal of optimizing the breaking time, a reference breaking time, a reference capacitance value, a reference resistance value, a reference decay time constant and a reference rated current are obtained;

[0191] Arrange the reference disconnection time, the reference capacitance value, the reference resistance value, the reference decay time constant and the reference rated current into a third training set and a third test set;

[0192] Preprocessing the third training set and the third test set;

[0193] Based on the preprocessed third training set, model training is performed to obtain a third neural network model;

[0194] Based on the preprocessed third test set, obtaining a third accuracy, a third precision, and a third recall rate of the third neural network model;

[0195] The third recall rate is used as a row vector, and the third accuracy rate is used as a column vector to form a ninth matrix, and then the reference disconnection time is introduced, and a surface fitting operation is performed to obtain a three-dimensional surface, and the three-dimensional surface is recorded as a ninth reference domain;

[0196] The third precision is used as a row vector, and the third accuracy rate is used as a column vector to form a tenth matrix, and then the reference disconnection time is introduced, and a surface fitting operation is performed to obtain a three-dimensional surface, and the three-dimensional surface is recorded as a tenth reference domain;

[0197] Based on the ninth reference domain and the tenth reference domain, with the reference disconnection time being the optimal target, using the third test set, selecting the optimal accuracy, optimal precision and optimal recall rate of the model corresponding to the optimal reference disconnection time;

[0198] Based on the optimal accuracy, the optimal precision and the optimal recall rate, the model parameters are optimized by adjusting the hyperparameters, modifying the model structure and adopting regularization, and based on the validation set, the model parameters corresponding to the optimal reference disconnection time are selected to obtain the optimal third neural network model;

[0199] The design of arc-free circuit breakers in the multi-PV access area is optimized based on the optimal third neural network model.

[0200] In some embodiments of this specification, the power supply voltage fluctuation amplitude of the multi-photovoltaic access lower stage area is obtained;

[0201] The power supply voltage fluctuation amplitude is used as a row vector, and the decay time constant is used as a column vector to form an eleventh matrix, and then the disconnection time is introduced, and a surface fitting operation is performed to obtain a three-dimensional surface and a power supply voltage fluctuation amplitude surface fitting equation, and the three-dimensional surface is recorded as an eleventh reference domain;

[0202] Based on the first reference domain, the second reference domain and the eleventh reference domain, with the goal of optimizing the disconnection time, a reference disconnection time, a reference capacitance value, a reference resistance value, a reference attenuation time constant and a power supply voltage fluctuation amplitude are obtained;

[0203] Arrange the reference disconnection time, the reference capacitance value, the reference resistance value, the reference decay time constant and the power supply voltage fluctuation amplitude into a fourth training set and a fourth test set;

[0204] Preprocessing the fourth training set and the fourth test set;

[0205] Based on the preprocessed fourth training set, model training is performed to obtain a fourth neural network model;

[0206] Based on the preprocessed fourth test set, obtaining a fourth accuracy, a fourth precision, and a fourth recall rate of the fourth neural network model;

[0207] The fourth recall rate is used as a row vector, and the fourth accuracy rate is used as a column vector to form a twelfth matrix, and then the reference disconnection time is introduced, and a surface fitting operation is performed to obtain a three-dimensional surface, and the three-dimensional surface is recorded as a twelfth reference domain;

[0208] The fourth precision is used as a row vector, the fourth accuracy rate is used as a column vector, a thirteenth matrix is ​​formed, and then the reference disconnection time is introduced, a surface fitting operation is performed, and a three-dimensional surface is obtained, and the three-dimensional surface is recorded as a thirteenth reference domain;

[0209] Based on the twelfth reference domain and the thirteenth reference domain, with the goal of optimizing the reference disconnection time, using the fourth test set, selecting the optimal accuracy, optimal precision and optimal recall rate of the model corresponding to the optimal reference disconnection time;

[0210] Based on the optimal accuracy, the optimal precision and the optimal recall rate, the model parameters are optimized by adjusting the hyperparameters, modifying the model structure and adopting regularization, and based on the validation set, the model parameters corresponding to the optimal reference disconnection time are selected to obtain the optimal fourth neural network model;

[0211] The design of arc-free circuit breakers in the multi-PV access area is optimized based on the optimal fourth neural network model.

[0212] In some embodiments of this specification, a current difference between a fault current of a multi-PV access lower section and a rated current of a circuit breaker circuit is obtained;

[0213] The current difference is used as a row vector and the decay time constant is used as a column vector to form a fourteenth matrix, and then the disconnection time is introduced to perform a surface fitting operation to obtain a three-dimensional surface and a current difference surface fitting equation, and the three-dimensional surface is recorded as a fourteenth reference domain;

[0214] Based on the first reference domain, the second reference domain and the fourteenth reference domain, with the goal of optimizing the disconnection time, a reference disconnection time, a reference capacitance value, a reference resistance value, a reference decay time constant and a reference current difference are obtained;

[0215] Arrange the reference disconnection time, the reference capacitance value, the reference resistance value, the reference decay time constant and the reference current difference value into a fifth training set and a fifth test set;

[0216] Preprocessing the fifth training set and the fifth test set;

[0217] Based on the preprocessed fifth training set, model training is performed to obtain a fifth neural network model;

[0218] Based on the preprocessed fifth test set, obtaining a fifth accuracy, a fifth precision, and a fifth recall rate of the fifth neural network model;

[0219] The fifth recall rate is used as a row vector, and the fifth accuracy rate is used as a column vector to form a fifteenth matrix, and then the reference disconnection time is imported, and a surface fitting operation is performed to obtain a three-dimensional surface, and the three-dimensional surface is recorded as a fifteenth reference domain;

[0220] The fifth precision is used as a row vector, the fifth accuracy rate is used as a column vector, a sixteenth matrix is ​​formed, and then the reference disconnection time is introduced, a surface fitting operation is performed, and a three-dimensional surface is obtained, and the three-dimensional surface is recorded as a sixteenth reference domain;

[0221] Based on the fifteenth reference domain and the sixteenth reference domain, with the goal of optimizing the reference disconnection time, using the fifth test set, selecting the optimal accuracy, optimal precision, and optimal recall rate of the model corresponding to the optimal reference disconnection time;

[0222] Based on the optimal accuracy, the optimal precision and the optimal recall rate, the model parameters are optimized by adjusting the hyperparameters, modifying the model structure and adopting regularization, and based on the validation set, the model parameters corresponding to the optimal reference disconnection time are selected to obtain the optimal fifth neural network model;

[0223] The design optimization of arc-free circuit breakers in the multi-PV access area is carried out based on the optimal fifth neural network model.

[0224] In some embodiments of this specification, obtaining an ambient temperature of a circuit breaker circuit;

[0225] The ambient temperature is used as a row vector and the decay time constant is used as a column vector to form a seventeenth matrix, and then the disconnection time is introduced to perform a surface fitting operation to obtain a three-dimensional surface and an ambient temperature surface fitting equation, and the three-dimensional surface is recorded as the seventeenth reference domain;

[0226] Based on the first reference domain, the second reference domain and the seventeenth reference domain, with the goal of optimizing the disconnection time, a reference disconnection time, a reference capacitance value, a reference resistance value, a reference decay time constant and a reference ambient temperature are obtained;

[0227] Arrange the reference disconnection time, the reference capacitance value, the reference resistance value, the reference decay time constant and the reference ambient temperature into a sixth training set and a sixth test set;

[0228] Preprocessing the sixth training set and the sixth test set;

[0229] Based on the preprocessed sixth training set, model training is performed to obtain a sixth neural network model;

[0230] Based on the preprocessed sixth test set, obtaining a sixth accuracy, a sixth precision, and a sixth recall rate of the sixth neural network model;

[0231] The sixth recall rate is used as a row vector, and the sixth accuracy rate is used as a column vector to form an eighteenth matrix, and then the reference disconnection time is introduced, and a surface fitting operation is performed to obtain a three-dimensional surface, and the three-dimensional surface is recorded as the eighteenth reference domain;

[0232] The sixth precision is used as a row vector, the sixth accuracy rate is used as a column vector, a nineteenth matrix is ​​formed, and then the reference disconnection time is introduced, a surface fitting operation is performed, and a three-dimensional surface is obtained, and the three-dimensional surface is recorded as a nineteenth reference domain;

[0233] Based on the eighteenth reference domain and the nineteenth reference domain, with the reference disconnection time being the optimal target, using the sixth test set, selecting the optimal accuracy, optimal precision and optimal recall rate of the model corresponding to the optimal reference disconnection time;

[0234] Based on the optimal accuracy, the optimal precision and the optimal recall rate, the model parameters are optimized by adjusting the hyperparameters, modifying the model structure and adopting regularization, and based on the validation set, the model parameters corresponding to the optimal reference disconnection time are selected to obtain the optimal sixth neural network model;

[0235] The design of arc-free circuit breakers in the multi-PV access area is optimized based on the optimal sixth neural network model.

[0236] In some embodiments of this specification, the harmonic current of the lower area where multiple photovoltaics are connected is obtained;

[0237] The harmonic current is used as a row vector and the decay time constant is used as a column vector to form a twentieth matrix, and then the disconnection time is introduced to perform a surface fitting operation to obtain a three-dimensional surface and a harmonic current surface fitting equation, and the three-dimensional surface is recorded as a twentieth reference domain;

[0238] Based on the first reference domain, the second reference domain and the twentieth reference domain, with the goal of optimizing the breaking time, a reference breaking time, a reference capacitance value, a reference resistance value, a reference attenuation time constant and a reference harmonic current are obtained;

[0239] Arrange the reference breaking time, the reference capacitance value, the reference resistance value, the reference decay time constant and the reference harmonic current into a seventh training set and a seventh test set;

[0240] Preprocessing the seventh training set and the seventh test set;

[0241] Based on the preprocessed seventh training set, model training is performed to obtain a seventh neural network model;

[0242] Based on the preprocessed seventh test set, obtaining a seventh accuracy, a seventh precision, and a seventh recall rate of the seventh neural network model;

[0243] The seventh recall rate is used as a row vector, and the seventh accuracy rate is used as a column vector to form a twenty-first matrix, and then the reference disconnection time is imported, and a surface fitting operation is performed to obtain a three-dimensional surface, and the three-dimensional surface is recorded as a twenty-first reference domain;

[0244] The seventh precision is used as a row vector, and the seventh accuracy rate is used as a column vector to form a twenty-second matrix, and then the reference disconnection time is introduced, and a surface fitting operation is performed to obtain a three-dimensional surface, and the three-dimensional surface is recorded as a twenty-second reference domain;

[0245] Based on the twenty-first reference domain and the twenty-second reference domain, with the reference disconnection time being the optimal target, using the seventh test set, selecting the optimal accuracy, optimal precision, and optimal recall rate of the model corresponding to the optimal reference disconnection time;

[0246] Based on the optimal accuracy, the optimal precision and the optimal recall rate, the model parameters are optimized by adjusting the hyperparameters, modifying the model structure and adopting regularization, and based on the validation set, the model parameters corresponding to the optimal reference disconnection time are selected to obtain the optimal seventh neural network model;

[0247] The design optimization of arc-free circuit breakers in the downstation area with multiple photovoltaic accesses is carried out based on the optimal seventh neural network model.

[0248] In some embodiments of this specification, the harmonic voltage of the lower area where multiple photovoltaics are connected is obtained;

[0249] The harmonic voltage is used as a row vector and the decay time constant is used as a column vector to form a twenty-third matrix, and then the disconnection time is introduced to perform a surface fitting operation to obtain a three-dimensional surface and a harmonic voltage surface fitting equation, and the three-dimensional surface is recorded as a twenty-third reference domain;

[0250] Based on the first reference domain, the second reference domain and the twenty-third reference domain, with the goal of optimizing the breaking time, a reference breaking time, a reference capacitance value, a reference resistance value, a reference attenuation time constant and a reference harmonic voltage are obtained;

[0251] Arrange the reference disconnection time, the reference capacitance value, the reference resistance value, the reference decay time constant and the reference harmonic voltage into an eighth training set and an eighth test set;

[0252] Preprocessing the eighth training set and the eighth test set;

[0253] Based on the preprocessed eighth training set, model training is performed to obtain an eighth neural network model;

[0254] Based on the preprocessed eighth test set, obtaining an eighth accuracy, an eighth precision, and an eighth recall rate of the eighth neural network model;

[0255] The eighth recall rate is used as a row vector, and the eighth accuracy rate is used as a column vector to form a twenty-fourth matrix, and then the reference disconnection time is imported, and a surface fitting operation is performed to obtain a three-dimensional surface, and the three-dimensional surface is recorded as a twenty-fourth reference domain;

[0256] The eighth precision is used as a row vector, the eighth accuracy rate is used as a column vector, a twenty-fifth matrix is ​​formed, and then the reference disconnection time is introduced, a surface fitting operation is performed, and a three-dimensional surface is obtained, and the three-dimensional surface is recorded as a twenty-fifth reference domain;

[0257] Based on the twenty-fourth reference domain and the twenty-fifth reference domain, with the reference disconnection time being the optimal target, using the eighth test set, selecting the optimal accuracy, optimal precision, and optimal recall rate of the model corresponding to the optimal reference disconnection time;

[0258] Based on the optimal accuracy, the optimal precision and the optimal recall rate, the model parameters are optimized by adjusting the hyperparameters, modifying the model structure and adopting regularization, and based on the validation set, the model parameters corresponding to the optimal reference disconnection time are selected to obtain the optimal eighth neural network model;

[0259] The design of arc-free circuit breakers in the multi-PV access area is optimized based on the optimal eighth neural network model.

[0260] In some embodiments of this specification, the harmonic frequency of the multiple photovoltaic access lower stage area is obtained;

[0261] The harmonic frequency is used as a row vector and the decay time constant is used as a column vector to form a twenty-sixth matrix, and then the disconnection time is introduced to perform a surface fitting operation to obtain a three-dimensional surface and a harmonic frequency surface fitting equation, and the three-dimensional surface is recorded as the twenty-sixth reference domain;

[0262] Based on the first reference domain, the second reference domain and the twenty-sixth reference domain, with the goal of optimizing the disconnection time, a reference disconnection time, a reference capacitance value, a reference resistance value, a reference attenuation time constant and a reference harmonic frequency are obtained;

[0263] Arrange the reference disconnection time, reference capacitance value, reference resistance value, reference decay time constant and reference harmonic frequency into a ninth training set and a ninth test set;

[0264] Preprocessing the ninth training set and the ninth test set;

[0265] Based on the preprocessed ninth training set, model training is performed to obtain a ninth neural network model;

[0266] Based on the preprocessed ninth test set, obtaining a ninth accuracy, a ninth precision, and a ninth recall rate of the ninth neural network model;

[0267] The ninth recall rate is used as a row vector, and the ninth accuracy rate is used as a column vector to form a twenty-seventh matrix, and then the reference disconnection time is imported, and a surface fitting operation is performed to obtain a three-dimensional surface, and the three-dimensional surface is recorded as a twenty-seventh reference domain;

[0268] The ninth precision is used as a row vector, the ninth accuracy rate is used as a column vector, a twenty-eighth matrix is ​​formed, and then the reference disconnection time is introduced, a surface fitting operation is performed, and a three-dimensional surface is obtained, and the three-dimensional surface is recorded as a twenty-eighth reference domain;

[0269] Based on the twenty-seventh reference domain and the twenty-eighth reference domain, with the reference disconnection time being the optimal target, using the ninth test set, selecting the optimal accuracy, optimal precision, and optimal recall rate of the model corresponding to the optimal reference disconnection time;

[0270] Based on the optimal accuracy, the optimal precision and the optimal recall rate, the model parameters are optimized by adjusting the hyperparameters, modifying the model structure and adopting regularization, and based on the validation set, the model parameters corresponding to the optimal reference disconnection time are selected to obtain the optimal ninth neural network model;

[0271] The design of arc-free circuit breakers in the multi-PV access substation is optimized based on the optimal ninth neural network model.

[0272] The embodiment of this specification also discloses a system for optimizing the design of arc-free circuit breakers in a multi-photovoltaic access lower area, which is used to implement any one of the above-mentioned methods for optimizing the design of arc-free circuit breakers in a multi-photovoltaic access lower area. The system for optimizing the design of arc-free circuit breakers in a multi-photovoltaic access lower area includes:

[0273] An acquisition module, used for acquiring a capacitance value, a resistance value, a decay time constant and a breaking time of a circuit breaker circuit;

[0274] The first fitting module is used to:

[0275] The capacitance value is used as a row vector, and the decay time constant is used as a column vector to form a first matrix, and then the disconnection time is introduced, and a surface fitting operation is performed to obtain a three-dimensional surface and a capacitance value surface fitting equation, and the three-dimensional surface is recorded as a first reference domain;

[0276] The resistance value is used as a row vector and the decay time constant is used as a column vector to form a second matrix, and then the disconnection time is introduced to perform a surface fitting operation to obtain a three-dimensional surface and a resistance value surface fitting equation, and the three-dimensional surface is recorded as a second reference domain;

[0277] A reference module, for obtaining a reference breaking time, a reference capacitance value, a reference resistance value, and a reference attenuation time constant based on the first reference domain and the second reference domain with the goal of optimizing the breaking time;

[0278] Neural network modules for:

[0279] Arrange the reference disconnection time, the reference capacitance value, the reference resistance value, and the reference decay time constant into a first training set and a first test set, and use the reference disconnection time alone as a verification set;

[0280] Preprocessing the first training set and the first test set;

[0281] Based on the preprocessed first training set, model training is performed to obtain a first neural network model;

[0282] Based on the preprocessed first test set, obtaining a first accuracy, a first precision, and a first recall of the first neural network model;

[0283] The second fitting module is used to:

[0284] The first recall rate is used as a row vector, and the first accuracy rate is used as a column vector to form a third matrix, and then the reference disconnection time is imported, and a surface fitting operation is performed to obtain a three-dimensional surface, and the three-dimensional surface is recorded as a third reference domain;

[0285] The first precision is used as a row vector, the first accuracy rate is used as a column vector, a fourth matrix is ​​formed, the reference disconnection time is introduced, a surface fitting operation is performed, a three-dimensional surface is obtained, and the three-dimensional surface is recorded as a fourth reference domain;

[0286] A first optimization module is used to select the optimal accuracy, optimal precision and optimal recall rate of the model corresponding to the optimal reference breaking time based on the third reference domain and the fourth reference domain and taking the optimal reference breaking time as the goal, using the first test set;

[0287] A second optimization module is used to optimize the model parameters by adjusting the hyperparameters, modifying the model structure and adopting regularization based on the optimal accuracy, the optimal precision and the optimal recall rate, and select the model parameters corresponding to the optimal reference disconnection time based on the validation set to obtain the optimal first neural network model;

[0288] The design optimization module is used to optimize the design of arc-free circuit breakers in the multi-PV access downstation area based on the optimal first neural network model.

[0289] It is clear that the above-mentioned embodiments can be used in combination, such as: using the optimal first neural network model to the optimal ninth neural network model (with the mean of the output) to jointly optimize the design of arc-free circuit breakers in the multi-photovoltaic access lower area (the optimization goal is to optimize the circuit breaking time, that is, the shortest time).

[0290] For example: first use the breaking time, attenuation time constant, capacitance value, resistance value, rated voltage, rated current, power supply voltage fluctuation amplitude, current difference, ambient temperature, harmonic current, harmonic voltage and harmonic frequency to perform surface fitting, and obtain the corresponding reference domains respectively, and then take the optimal breaking time as the goal to obtain the corresponding reference data, and perform neural network training based on these reference data to obtain the corresponding neural network model, and perform surface fitting based on the accuracy, precision and recall rate of these neural network models to obtain the corresponding reference domains respectively, and then take the optimal breaking time as the goal to obtain the optimal accuracy, optimal precision and optimal recall rate of the model corresponding to the optimal reference breaking time, based on the optimal accuracy, optimal precision and optimal recall rate, optimize the model parameters by adjusting the hyperparameters, modifying the model structure and using regularization, and select the model parameters corresponding to the optimal reference breaking time based on the validation set to obtain the optimal neural network model, and finally optimize the breaking time of the photovoltaic inverter based on the optimal neural network model.

[0291] The present invention is mainly based on the optimal neural network model to change and adjust the electrical parameters of the circuit breaker circuit itself and the electrical parameters of the photovoltaic system to perform design optimization. As for other design parameters of the arc-free circuit breaker, such as mechanical structure, arc-free technology, topological structure, control strategy, etc., reference can be made to the prior art. The focus of the present invention is: based on the optimal neural network model, the above-mentioned electrical parameters are changed to optimize the circuit-breaking time, and based on this design optimization idea, a suitable and optimal arc-free circuit breaker to be applied to the current multi-photovoltaic access downstage area is designed.

[0292] The present invention adopts a surface fitting method, which can well reflect the influence of the attenuation time constant, capacitance value, resistance value, rated voltage, rated current, power supply voltage fluctuation amplitude, current difference, ambient temperature, harmonic current, harmonic voltage and harmonic frequency on the circuit breaking time, as well as the influence of the attenuation time constant, capacitance value, resistance value, rated voltage, rated current, power supply voltage fluctuation amplitude, current difference, ambient temperature, harmonic current, harmonic voltage and harmonic frequency on the circuit breaking time in different combinations.

[0293] It is understandable that for each optimization, these data and their corresponding circuit-breaking times can be obtained and then used to update the reference domain. Through long-term updates, the optimization design effect and accuracy can be improved.

[0294] Definition explanation:

[0295] Decay time constant: The time required for the residual current in the circuit to decay to a specified value after the circuit breaker cuts off the current. A shorter decay time constant helps to reduce the duration of the arc and improve the performance of the circuit breaker.

[0296] Breaking time: The time from triggering the circuit breaker to applying the cut-off current. The breaking time of the arc-free circuit breaker should be as short as possible to reduce the generation of arcs and extend the life of electrical equipment.

[0297] Rated current and fault current: When a fault occurs in the photovoltaic system, such as a short circuit or overload, the fault current generated will trigger the arcless circuit breaker to operate. The size of the fault current will directly affect the circuit-breaking time of the arcless circuit breaker. Usually, in a photovoltaic system, when the current exceeds the rated current of the circuit breaker by several times (such as 10 times or 12 times), the circuit breaker will operate quickly. Since the fault current is difficult to predict, it cannot be adjusted directly, but the current difference can be adjusted by adjusting the rated current; that is, the present invention is based on the historical data or simulation data of the fault current in the photovoltaic system, and adopts a surface fitting method, with the goal of optimizing the circuit-breaking time, to determine the size of the rated current of the circuit breaker currently required (to be designed) (that is, determine the design requirements).

[0298] Harmonic current, harmonic voltage and harmonic frequency: Harmonics refer to the components of current or voltage whose frequency is an integer multiple of the fundamental frequency. In photovoltaic power generation systems, harmonics are often generated due to the nonlinear characteristics of inverters and other electrical equipment. Harmonics can cause voltage and current distortion in photovoltaic systems. The operation of arc-free circuit breakers in photovoltaic systems may be affected by the control system, including the generation, transmission and processing of trigger signals. For example, harmonic currents, harmonic voltages and harmonic frequencies will directly and / or indirectly affect the operation of arc-free circuit breakers.

[0299] It is understandable that corresponding functional modules may be added when implementing the methods of the above embodiments.

[0300] In practice, before surface fitting, the matrix can be preprocessed to remove NaN values ​​or 0 values, and then the break time can be imported to perform the surface fitting operation.

[0301] It should be understood that the above-mentioned surface fitting operation can use existing software tools, such as the Curve Fitting tool in Matlab. The above-mentioned modules can be data reading modules of the existing data detection (collection) / measurement device / sensor or the arc-free circuit breaker design optimization system for multiple photovoltaic access downstage areas, and can directly obtain corresponding data.

[0302] It is clear that the corresponding data in the arc-free circuit breaker design optimization system for multiple photovoltaic access downstage areas can be obtained in advance through existing equipment, that is, a large amount of test data can be obtained. These test data can be simulation experiment data in the laboratory, or actual measured or operating data of the arc-free circuit breaker design optimization system for multiple photovoltaic access downstage areas in practice.

[0303] The above-described embodiments are used to illustrate the present invention, not to limit the present invention, so changes in the exemplified values ​​or replacement of equivalent elements should still fall within the scope of the present invention.

[0304] From the above detailed description, it can be understood by those skilled in the art that the present invention can indeed achieve the aforementioned objectives and is in compliance with the provisions of the Patent Law.

[0305] Although the preferred embodiments of the present invention have been described, those skilled in the art may make additional changes and modifications to these embodiments once they are aware of the basic creative concepts. 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. The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. It should be pointed out that any modifications, equivalent substitutions and improvements made within the spirit and principles of the present invention should be included in the scope of protection of the present invention.

[0306] It should be noted that the above description of the relevant process is only for example and explanation, and does not limit the scope of application of this specification. For those skilled in the art, various modifications and changes can be made to the process under the guidance of this specification. However, these modifications and changes are still within the scope of this specification.

[0307] The basic concepts have been described above. Obviously, for those of ordinary skill in the art who have read this application, the above invention disclosure is only for example and does not constitute a limitation of this application. Although not explicitly stated here, those of ordinary skill in the art may make various modifications, improvements and amendments to this application. Such modifications, improvements and amendments are suggested in this application, so such modifications, improvements and amendments still belong to the spirit and scope of the exemplary embodiments of this application.

[0308] At the same time, the present application uses specific words to describe the embodiments of the present application. For example, "one embodiment", "an embodiment", and / or "some embodiments" refer to a certain feature, structure or characteristic related to at least one embodiment of the present application. Therefore, it should be emphasized and noted that "one embodiment" or "an embodiment" or "an alternative embodiment" mentioned twice or more in different positions in this specification does not necessarily refer to the same embodiment. In addition, some features, structures or characteristics in one or more embodiments of the present application can be appropriately combined.

[0309] In addition, it will be appreciated by those skilled in the art that various aspects of the present application may be illustrated and described by a number of patentable categories or situations, including any new and useful combination of processes, machines, products or substances, or any new and useful improvements thereto. Therefore, various aspects of the present application may be implemented entirely by hardware, entirely by software (including firmware, resident software, microcode, etc.), or by a combination of hardware and software. The above hardware or software may all be referred to as "units", "modules" or "systems". In addition, various aspects of the present application may take the form of a computer program product embodied in one or more computer-readable media, wherein computer-readable program code is contained therein.

[0310] In addition, unless explicitly stated in the claims, the order of the processing elements and sequences described in this application, the use of alphanumeric characters, or the use of other names are not intended to limit the order of the processes and methods of this application. Although the above disclosure discusses some invention embodiments that are currently considered useful through various examples, it should be understood that such details are only for illustrative purposes, and the attached claims are not limited to the disclosed embodiments. On the contrary, the claims are intended to cover all modifications and equivalent combinations that are consistent with the essence and scope of the embodiments of this application. For example, although the implementation of the various components described above can be embodied in a hardware device, it can also be implemented as a pure software solution, for example, installation on an existing server or mobile device.

[0311] Similarly, it should be noted that in order to simplify the description disclosed in this application and thus help understand one or more embodiments of the invention, in the above description of the embodiments of the application, multiple features are sometimes combined into one embodiment, figure or description thereof. However, this method of the application should not be interpreted as reflecting the intention that the claimed object requires more features than those explicitly stated in each claim. On the contrary, the subject of the invention should have fewer features than the above single embodiment.

Claims

1. A method for optimizing the design of arc-free circuit breakers for multiple photovoltaic access areas, characterized in that: include: Obtain the capacitance value, resistance value, decay time constant and breaking time of the circuit breaker circuit; The capacitance value is used as a row vector and the decay time constant is used as a column vector to form a first matrix, and then the disconnection time is introduced to perform a surface fitting operation to obtain a three-dimensional surface and a capacitance value surface fitting equation, and the three-dimensional surface is recorded as a first reference domain; The resistance value is used as a row vector and the decay time constant is used as a column vector to form a second matrix, and then the disconnection time is introduced to perform a surface fitting operation to obtain a three-dimensional surface and a resistance value surface fitting equation, and the three-dimensional surface is recorded as a second reference domain; Based on the first reference domain and the second reference domain, with the goal of optimizing the disconnection time, a reference disconnection time, a reference capacitance value, a reference resistance value, and a reference decay time constant are obtained; Arrange the reference disconnection time, the reference capacitance value, the reference resistance value, and the reference decay time constant into a first training set and a first test set, and use the reference disconnection time alone as a verification set; Preprocessing the first training set and the first test set; Based on the preprocessed first training set, model training is performed to obtain a first neural network model; Based on the preprocessed first test set, obtaining a first accuracy, a first precision, and a first recall of the first neural network model; The first recall rate is used as a row vector, and the first accuracy rate is used as a column vector to form a third matrix, and then the reference disconnection time is imported, and a surface fitting operation is performed to obtain a three-dimensional surface, and the three-dimensional surface is recorded as a third reference domain; The first precision is used as a row vector, the first accuracy rate is used as a column vector, a fourth matrix is ​​formed, the reference disconnection time is introduced, a surface fitting operation is performed, a three-dimensional surface is obtained, and the three-dimensional surface is recorded as a fourth reference domain; Based on the third reference domain and the fourth reference domain, with the goal of optimizing the reference disconnection time, using the first test set, selecting the optimal accuracy, optimal precision, and optimal recall rate of the model corresponding to the optimal reference disconnection time; Based on the optimal accuracy, the optimal precision and the optimal recall rate, the model parameters are optimized by adjusting the hyperparameters, modifying the model structure and using regularization, and based on the validation set, the model parameters corresponding to the optimal reference disconnection time are selected to obtain the optimal first neural network model; The design optimization of arc-free circuit breakers in the multi-PV access area is carried out based on the optimal first neural network model.

2. The arc-free circuit breaker design optimization method for multiple photovoltaic access downstations according to claim 1 is characterized in that: Get the rated voltage of the circuit breaker circuit; The rated voltage is used as a row vector and the decay time constant is used as a column vector to form a fifth matrix, and then the disconnection time is introduced to perform a surface fitting operation to obtain a three-dimensional surface and a rated voltage surface fitting equation, and the three-dimensional surface is recorded as a fifth reference domain; Based on the first reference domain, the second reference domain and the fifth reference domain, with the goal of optimizing the breaking time, a reference breaking time, a reference capacitance value, a reference resistance value, a reference decay time constant and a reference rated voltage are obtained; Arrange the reference disconnection time, the reference capacitance value, the reference resistance value, the reference decay time constant and the reference rated voltage into a second training set and a second test set; Preprocessing the second training set and the second test set; Based on the preprocessed second training set, model training is performed to obtain a second neural network model; Based on the preprocessed second test set, obtaining a second accuracy, a second precision, and a second recall of the second neural network model; The second recall rate is used as a row vector, and the second accuracy rate is used as a column vector to form a sixth matrix, and then the reference disconnection time is imported, and a surface fitting operation is performed to obtain a three-dimensional surface, and the three-dimensional surface is recorded as a sixth reference domain; The second precision is used as a row vector, and the second accuracy rate is used as a column vector to form a seventh matrix, and then the reference disconnection time is introduced, and a surface fitting operation is performed to obtain a three-dimensional surface, and the three-dimensional surface is recorded as a seventh reference domain; Based on the sixth reference domain and the seventh reference domain, with the goal of optimizing the reference disconnection time, using the second test set, selecting the optimal accuracy, optimal precision, and optimal recall rate of the model corresponding to the optimal reference disconnection time; Based on the optimal accuracy, the optimal precision and the optimal recall rate, the model parameters are optimized by adjusting the hyperparameters, modifying the model structure and using regularization, and based on the validation set, the model parameters corresponding to the optimal reference disconnection time are selected to obtain the optimal second neural network model; The design of arc-free circuit breakers in the multi-PV access area is optimized based on the optimal second neural network model.

3. The arc-free circuit breaker design optimization method for multiple photovoltaic access downstations according to claim 1 is characterized in that: Get the rated current of the circuit breaker circuit; The rated current is used as a row vector and the decay time constant is used as a column vector to form an eighth matrix, and then the disconnection time is introduced to perform a surface fitting operation to obtain a three-dimensional surface and a rated current surface fitting equation, and the three-dimensional surface is recorded as an eighth reference domain; Based on the first reference domain, the second reference domain and the eighth reference domain, with the goal of optimizing the breaking time, a reference breaking time, a reference capacitance value, a reference resistance value, a reference decay time constant and a reference rated current are obtained; Arrange the reference disconnection time, the reference capacitance value, the reference resistance value, the reference decay time constant and the reference rated current into a third training set and a third test set; Preprocessing the third training set and the third test set; Based on the preprocessed third training set, model training is performed to obtain a third neural network model; Based on the preprocessed third test set, obtaining a third accuracy, a third precision, and a third recall rate of the third neural network model; The third recall rate is used as a row vector, and the third accuracy rate is used as a column vector to form a ninth matrix, and then the reference disconnection time is introduced, and a surface fitting operation is performed to obtain a three-dimensional surface, and the three-dimensional surface is recorded as a ninth reference domain; The third precision is used as a row vector, and the third accuracy rate is used as a column vector to form a tenth matrix, and then the reference disconnection time is introduced, and a surface fitting operation is performed to obtain a three-dimensional surface, and the three-dimensional surface is recorded as a tenth reference domain; Based on the ninth reference domain and the tenth reference domain, with the reference disconnection time being the optimal target, using the third test set, selecting the optimal accuracy, optimal precision and optimal recall rate of the model corresponding to the optimal reference disconnection time; Based on the optimal accuracy, the optimal precision and the optimal recall rate, the model parameters are optimized by adjusting the hyperparameters, modifying the model structure and adopting regularization, and based on the validation set, the model parameters corresponding to the optimal reference disconnection time are selected to obtain the optimal third neural network model; The design of arc-free circuit breakers in the multi-PV access area is optimized based on the optimal third neural network model.

4. The arc-free circuit breaker design optimization method for multiple photovoltaic access downstations according to claim 1 is characterized in that: Obtain the power supply voltage fluctuation amplitude of the multi-PV access area; The power supply voltage fluctuation amplitude is used as a row vector, and the decay time constant is used as a column vector to form an eleventh matrix, and then the disconnection time is introduced, and a surface fitting operation is performed to obtain a three-dimensional surface and a power supply voltage fluctuation amplitude surface fitting equation, and the three-dimensional surface is recorded as an eleventh reference domain; Based on the first reference domain, the second reference domain and the eleventh reference domain, with the goal of optimizing the disconnection time, a reference disconnection time, a reference capacitance value, a reference resistance value, a reference attenuation time constant and a power supply voltage fluctuation amplitude are obtained; Arrange the reference disconnection time, the reference capacitance value, the reference resistance value, the reference decay time constant and the power supply voltage fluctuation amplitude into a fourth training set and a fourth test set; Preprocessing the fourth training set and the fourth test set; Based on the preprocessed fourth training set, model training is performed to obtain a fourth neural network model; Based on the preprocessed fourth test set, obtaining a fourth accuracy, a fourth precision, and a fourth recall rate of the fourth neural network model; The fourth recall rate is used as a row vector, and the fourth accuracy rate is used as a column vector to form a twelfth matrix, and then the reference disconnection time is introduced, and a surface fitting operation is performed to obtain a three-dimensional surface, and the three-dimensional surface is recorded as a twelfth reference domain; The fourth precision is used as a row vector, the fourth accuracy rate is used as a column vector, a thirteenth matrix is ​​formed, and then the reference disconnection time is introduced, a surface fitting operation is performed, and a three-dimensional surface is obtained, and the three-dimensional surface is recorded as a thirteenth reference domain; Based on the twelfth reference domain and the thirteenth reference domain, with the goal of optimizing the reference disconnection time, using the fourth test set, selecting the optimal accuracy, optimal precision and optimal recall rate of the model corresponding to the optimal reference disconnection time; Based on the optimal accuracy, optimal precision and optimal recall rate, the model parameters are optimized by adjusting the hyperparameters, modifying the model structure and using regularization, and based on the verification set, the model parameters corresponding to the optimal reference circuit breaking time are selected to obtain the optimal fourth neural network model; based on the optimal fourth neural network model, the design of arc-free circuit breakers in the lower-level area of ​​multi-photovoltaic access is optimized.

5. The arc-free circuit breaker design optimization method for multiple photovoltaic access downstations according to claim 1 is characterized in that: Obtain the current difference between the fault current of the multi-PV access lower section and the rated current of the circuit breaker circuit; The current difference is used as a row vector and the decay time constant is used as a column vector to form a fourteenth matrix, and then the disconnection time is introduced to perform a surface fitting operation to obtain a three-dimensional surface and a current difference surface fitting equation, and the three-dimensional surface is recorded as a fourteenth reference domain; Based on the first reference domain, the second reference domain and the fourteenth reference domain, with the goal of optimizing the disconnection time, a reference disconnection time, a reference capacitance value, a reference resistance value, a reference decay time constant and a reference current difference are obtained; Arrange the reference disconnection time, the reference capacitance value, the reference resistance value, the reference decay time constant and the reference current difference value into a fifth training set and a fifth test set; Preprocessing the fifth training set and the fifth test set; Based on the preprocessed fifth training set, model training is performed to obtain a fifth neural network model; Based on the preprocessed fifth test set, obtaining a fifth accuracy, a fifth precision, and a fifth recall rate of the fifth neural network model; The fifth recall rate is used as a row vector, and the fifth accuracy rate is used as a column vector to form a fifteenth matrix, and then the reference disconnection time is imported, and a surface fitting operation is performed to obtain a three-dimensional surface, and the three-dimensional surface is recorded as a fifteenth reference domain; The fifth precision is used as a row vector, the fifth accuracy rate is used as a column vector, a sixteenth matrix is ​​formed, and then the reference disconnection time is introduced, a surface fitting operation is performed, and a three-dimensional surface is obtained, and the three-dimensional surface is recorded as a sixteenth reference domain; Based on the fifteenth reference domain and the sixteenth reference domain, with the goal of optimizing the reference disconnection time, using the fifth test set, selecting the optimal accuracy, optimal precision, and optimal recall rate of the model corresponding to the optimal reference disconnection time; Based on the optimal accuracy, the optimal precision and the optimal recall rate, the model parameters are optimized by adjusting the hyperparameters, modifying the model structure and adopting regularization, and based on the validation set, the model parameters corresponding to the optimal reference disconnection time are selected to obtain the optimal fifth neural network model; The design optimization of arc-free circuit breakers in the multi-PV access substation is carried out based on the optimal fifth neural network model.

6. The arc-free circuit breaker design optimization method for multiple photovoltaic access downstations according to claim 1 is characterized in that: Get the ambient temperature of the circuit breaker circuit; The ambient temperature is used as a row vector and the decay time constant is used as a column vector to form a seventeenth matrix, and then the disconnection time is introduced to perform a surface fitting operation to obtain a three-dimensional surface and an ambient temperature surface fitting equation, and the three-dimensional surface is recorded as the seventeenth reference domain; Based on the first reference domain, the second reference domain and the seventeenth reference domain, with the goal of optimizing the disconnection time, a reference disconnection time, a reference capacitance value, a reference resistance value, a reference decay time constant and a reference ambient temperature are obtained; Arrange the reference disconnection time, the reference capacitance value, the reference resistance value, the reference decay time constant and the reference ambient temperature into a sixth training set and a sixth test set; Preprocessing the sixth training set and the sixth test set; Based on the preprocessed sixth training set, model training is performed to obtain a sixth neural network model; Based on the preprocessed sixth test set, obtaining a sixth accuracy, a sixth precision, and a sixth recall rate of the sixth neural network model; The sixth recall rate is used as a row vector, and the sixth accuracy rate is used as a column vector to form an eighteenth matrix, and then the reference disconnection time is introduced, and a surface fitting operation is performed to obtain a three-dimensional surface, and the three-dimensional surface is recorded as the eighteenth reference domain; The sixth precision is used as a row vector, the sixth accuracy rate is used as a column vector, a nineteenth matrix is ​​formed, and then the reference disconnection time is introduced, a surface fitting operation is performed, and a three-dimensional surface is obtained, and the three-dimensional surface is recorded as a nineteenth reference domain; Based on the eighteenth reference domain and the nineteenth reference domain, with the reference disconnection time being the optimal target, using the sixth test set, selecting the optimal accuracy, optimal precision and optimal recall rate of the model corresponding to the optimal reference disconnection time; Based on the optimal accuracy, the optimal precision and the optimal recall rate, the model parameters are optimized by adjusting the hyperparameters, modifying the model structure and adopting regularization, and based on the validation set, the model parameters corresponding to the optimal reference disconnection time are selected to obtain the optimal sixth neural network model; The design of arc-free circuit breakers in the multi-PV access area is optimized based on the optimal sixth neural network model.

7. The arc-free circuit breaker design optimization method for multiple photovoltaic access downstations according to claim 1 is characterized in that: Obtain the harmonic current of multiple photovoltaic access areas; The harmonic current is used as a row vector and the decay time constant is used as a column vector to form a twentieth matrix, and then the disconnection time is introduced to perform a surface fitting operation to obtain a three-dimensional surface and a harmonic current surface fitting equation, and the three-dimensional surface is recorded as a twentieth reference domain; Based on the first reference domain, the second reference domain and the twentieth reference domain, with the goal of optimizing the breaking time, a reference breaking time, a reference capacitance value, a reference resistance value, a reference attenuation time constant and a reference harmonic current are obtained; Arrange the reference breaking time, the reference capacitance value, the reference resistance value, the reference decay time constant and the reference harmonic current into a seventh training set and a seventh test set; Preprocessing the seventh training set and the seventh test set; Based on the preprocessed seventh training set, model training is performed to obtain a seventh neural network model; Based on the preprocessed seventh test set, obtaining a seventh accuracy, a seventh precision, and a seventh recall rate of the seventh neural network model; The seventh recall rate is used as a row vector, and the seventh accuracy rate is used as a column vector to form a twenty-first matrix, and then the reference disconnection time is imported, and a surface fitting operation is performed to obtain a three-dimensional surface, and the three-dimensional surface is recorded as a twenty-first reference domain; The seventh precision is used as a row vector, and the seventh accuracy rate is used as a column vector to form a twenty-second matrix, and then the reference disconnection time is introduced, and a surface fitting operation is performed to obtain a three-dimensional surface, and the three-dimensional surface is recorded as a twenty-second reference domain; Based on the twenty-first reference domain and the twenty-second reference domain, with the reference disconnection time being the optimal target, using the seventh test set, selecting the optimal accuracy, optimal precision, and optimal recall rate of the model corresponding to the optimal reference disconnection time; Based on the optimal accuracy, the optimal precision and the optimal recall rate, the model parameters are optimized by adjusting the hyperparameters, modifying the model structure and adopting regularization, and based on the validation set, the model parameters corresponding to the optimal reference disconnection time are selected to obtain the optimal seventh neural network model; The design optimization of arc-free circuit breakers in the downstation area with multiple photovoltaic accesses is carried out based on the optimal seventh neural network model.

8. The arc-free circuit breaker design optimization method for multiple photovoltaic access downstations according to claim 1 is characterized in that: Obtain the harmonic voltage of the area where multiple photovoltaics are connected; The harmonic voltage is used as a row vector and the decay time constant is used as a column vector to form a twenty-third matrix, and then the disconnection time is introduced to perform a surface fitting operation to obtain a three-dimensional surface and a harmonic voltage surface fitting equation, and the three-dimensional surface is recorded as a twenty-third reference domain; Based on the first reference domain, the second reference domain and the twenty-third reference domain, with the goal of optimizing the breaking time, a reference breaking time, a reference capacitance value, a reference resistance value, a reference attenuation time constant and a reference harmonic voltage are obtained; Arrange the reference disconnection time, the reference capacitance value, the reference resistance value, the reference decay time constant and the reference harmonic voltage into an eighth training set and an eighth test set; Preprocessing the eighth training set and the eighth test set; Based on the preprocessed eighth training set, model training is performed to obtain an eighth neural network model; Based on the preprocessed eighth test set, obtaining an eighth accuracy, an eighth precision, and an eighth recall rate of the eighth neural network model; The eighth recall rate is used as a row vector, and the eighth accuracy rate is used as a column vector to form a twenty-fourth matrix, and then the reference disconnection time is imported, and a surface fitting operation is performed to obtain a three-dimensional surface, and the three-dimensional surface is recorded as a twenty-fourth reference domain; The eighth precision is used as a row vector, the eighth accuracy rate is used as a column vector, a twenty-fifth matrix is ​​formed, and then the reference disconnection time is introduced, a surface fitting operation is performed, and a three-dimensional surface is obtained, and the three-dimensional surface is recorded as a twenty-fifth reference domain; Based on the twenty-fourth reference domain and the twenty-fifth reference domain, with the reference disconnection time being the optimal target, using the eighth test set, selecting the optimal accuracy, optimal precision, and optimal recall rate of the model corresponding to the optimal reference disconnection time; Based on the optimal accuracy, the optimal precision and the optimal recall rate, the model parameters are optimized by adjusting the hyperparameters, modifying the model structure and adopting regularization, and based on the validation set, the model parameters corresponding to the optimal reference disconnection time are selected to obtain the optimal eighth neural network model; The design of arc-free circuit breakers in the multi-PV access area is optimized based on the optimal eighth neural network model.

9. The arc-free circuit breaker design optimization method for multiple photovoltaic access downstations according to claim 1 is characterized in that: Obtain the harmonic frequency of multiple photovoltaic access areas; The harmonic frequency is used as a row vector and the decay time constant is used as a column vector to form a twenty-sixth matrix, and then the disconnection time is introduced to perform a surface fitting operation to obtain a three-dimensional surface and a harmonic frequency surface fitting equation, and the three-dimensional surface is recorded as the twenty-sixth reference domain; Based on the first reference domain, the second reference domain and the twenty-sixth reference domain, with the goal of optimizing the disconnection time, a reference disconnection time, a reference capacitance value, a reference resistance value, a reference attenuation time constant and a reference harmonic frequency are obtained; Arrange the reference disconnection time, reference capacitance value, reference resistance value, reference decay time constant and reference harmonic frequency into a ninth training set and a ninth test set; Preprocessing the ninth training set and the ninth test set; Based on the preprocessed ninth training set, model training is performed to obtain a ninth neural network model; Based on the preprocessed ninth test set, obtaining a ninth accuracy, a ninth precision, and a ninth recall rate of the ninth neural network model; The ninth recall rate is used as a row vector, and the ninth accuracy rate is used as a column vector to form a twenty-seventh matrix, and then the reference disconnection time is imported, and a surface fitting operation is performed to obtain a three-dimensional surface, and the three-dimensional surface is recorded as a twenty-seventh reference domain; The ninth precision is used as a row vector, the ninth accuracy rate is used as a column vector, a twenty-eighth matrix is ​​formed, and then the reference disconnection time is introduced, a surface fitting operation is performed, and a three-dimensional surface is obtained, and the three-dimensional surface is recorded as a twenty-eighth reference domain; Based on the twenty-seventh reference domain and the twenty-eighth reference domain, with the reference disconnection time being the optimal target, using the ninth test set, selecting the optimal accuracy, optimal precision, and optimal recall rate of the model corresponding to the optimal reference disconnection time; Based on the optimal accuracy, the optimal precision and the optimal recall rate, the model parameters are optimized by adjusting the hyperparameters, modifying the model structure and adopting regularization, and based on the validation set, the model parameters corresponding to the optimal reference disconnection time are selected to obtain the optimal ninth neural network model; The design of arc-free circuit breakers in the multi-PV access substation is optimized based on the optimal ninth neural network model.

10. A design optimization system for arc-free circuit breakers in multiple photovoltaic access areas, characterized in that: The arc-free circuit breaker design optimization method for implementing the multi-photovoltaic access to the lower stage area according to any one of claims 1 to 9, the arc-free circuit breaker design optimization system for the multi-photovoltaic access to the lower stage area comprises: An acquisition module, used for acquiring a capacitance value, a resistance value, a decay time constant and a breaking time of a circuit breaker circuit; The first fitting module is used to: The capacitance value is used as a row vector, and the decay time constant is used as a column vector to form a first matrix, and then the disconnection time is introduced, and a surface fitting operation is performed to obtain a three-dimensional surface and a capacitance value surface fitting equation, and the three-dimensional surface is recorded as a first reference domain; The resistance value is used as a row vector and the decay time constant is used as a column vector to form a second matrix, and then the disconnection time is introduced to perform a surface fitting operation to obtain a three-dimensional surface and a resistance value surface fitting equation, and the three-dimensional surface is recorded as a second reference domain; A reference module, for obtaining a reference breaking time, a reference capacitance value, a reference resistance value, and a reference attenuation time constant based on the first reference domain and the second reference domain with the goal of optimizing the breaking time; Neural network modules for: Arrange the reference disconnection time, the reference capacitance value, the reference resistance value, and the reference decay time constant into a first training set and a first test set, and use the reference disconnection time alone as a verification set; Preprocessing the first training set and the first test set; Based on the preprocessed first training set, model training is performed to obtain a first neural network model; Based on the preprocessed first test set, obtaining a first accuracy, a first precision, and a first recall of the first neural network model; The second fitting module is used to: The first recall rate is used as a row vector, and the first accuracy rate is used as a column vector to form a third matrix, and then the reference disconnection time is imported, and a surface fitting operation is performed to obtain a three-dimensional surface, and the three-dimensional surface is recorded as a third reference domain; The first precision is used as a row vector, the first accuracy rate is used as a column vector, a fourth matrix is ​​formed, the reference disconnection time is introduced, a surface fitting operation is performed, a three-dimensional surface is obtained, and the three-dimensional surface is recorded as a fourth reference domain; A first optimization module is used to select the optimal accuracy, optimal precision and optimal recall rate of the model corresponding to the optimal reference breaking time based on the third reference domain and the fourth reference domain and taking the optimal reference breaking time as the goal, using the first test set; A second optimization module is used to optimize the model parameters by adjusting the hyperparameters, modifying the model structure and adopting regularization based on the optimal accuracy, the optimal precision and the optimal recall rate, and select the model parameters corresponding to the optimal reference disconnection time based on the validation set to obtain the optimal first neural network model; The design optimization module is used to optimize the design of arc-free circuit breakers in the multi-PV access downstation area based on the optimal first neural network model.

Images