Photovoltaic power processing method and system

Abstract

The present invention proposes a photovoltaic power processing method and system, which belongs to the field of photovoltaic microgrid control technology. The system includes a photovoltaic energy combination unit, multiple DC / DC converters, a capacitor unit, a battery unit and a capacitor state sensing unit; the capacitor state sensing unit is used to sense the charge state of the capacitor unit; the battery unit is charged or discharged based on the charge state sensed by the capacitor state sensing unit. The method includes when the output power value of the photovoltaic energy combination unit in the future target preset time period matches the current output power of the AC large power grid to be connected to the grid, the photovoltaic energy combination unit is connected to the AC large power grid to be connected to the grid. The present application can realize intelligent power distribution and scheduling of photovoltaic microgrids containing DC and AC mixed loads, enhance the robustness of microgrid power prediction, effectively improve the voltage output of photovoltaic power grid connection, and enhance the reliability of the distribution network system.

Description

Technical Field

[0001] The present invention belongs to the technical field of photovoltaic microgrid control, and in particular relates to a photovoltaic power processing method and system. Background Art

[0002] Solar photovoltaic power generation system refers to the use of solar cells to achieve the transformation of solar radiation and convert it into electrical energy. Solar photovoltaic power generation system is a new type of power generation system, which mainly transforms sunlight into electrical energy. Network operation and off-grid operation are the main operating modes of this system.

[0003] Since photovoltaic power generation is easily constrained by the natural environment and has the characteristics of instability, intermittency and volatility, it is usually equipped with energy storage devices such as supercapacitors to absorb excess electricity when there is excess electricity and provide electricity when there is insufficient electricity, thereby maintaining the stability of the power system. Photovoltaic power generation is connected to the main power grid in the form of microgrids, making the operation of the power system more complicated and increasing the randomness and uncertainty of the power system flow distribution.

[0004] With the increase in the capacity of photovoltaic power generation connected to the grid, the accuracy of photovoltaic microgrid output power prediction has affected the safe and stable operation of the power grid; at the same time, in DC or AC / DC hybrid bus microgrids, photovoltaic and other micro power sources and loads can be connected in parallel to the DC bus and directly exchange power with the energy storage system. How to achieve short-term high-rate charging and discharging tasks, thereby reducing the battery charging and discharging current and cycle times to extend the battery life, and achieve reliable microgrid hybrid load power distribution and safe grid-connected operation has also become a difficult problem faced in the process of photovoltaic microgrid automation control. Summary of the invention

[0005] In order to solve the above technical problems, the present invention provides a photovoltaic power processing method and system.

[0006] In a first aspect of the present invention, a photovoltaic power processing system is provided, the system comprising a photovoltaic energy combination unit, a plurality of DC / DC converters, a capacitor unit, a battery unit, and a capacitor state sensing unit;

[0007] Each of the photovoltaic energy combination units includes a target photovoltaic energy panel and a left adjacent photovoltaic energy panel and a right adjacent photovoltaic energy panel of the target photovoltaic energy panel;

[0008] The target photovoltaic energy panel is connected to the DC bus through a first DC / DC converter, the left neighboring photovoltaic energy panel of the target photovoltaic energy panel is connected to a second DC / DC converter, and the right neighboring photovoltaic energy panel of the target photovoltaic energy panel is connected to a third DC / DC converter;

[0009] The battery unit is connected to the target photovoltaic energy panel through a first DC / DC converter corresponding to the target photovoltaic energy panel, and is connected to the capacitor unit;

[0010] The left neighboring photovoltaic energy panel and the right neighboring photovoltaic energy panel of the target photovoltaic energy panel are connected to the battery unit through the capacitor unit;

[0011] The capacitor state sensing unit is used to sense the charge state of the capacitor unit;

[0012] The battery unit is charged or discharged based on the charge state sensed by the capacitance state sensing unit.

[0013] The capacitor unit includes a first super capacitor and a second super capacitor;

[0014] The left neighboring photovoltaic energy panel is connected to the first supercapacitor via the second DC / DC converter;

[0015] The right-adjacent photovoltaic energy panel is connected to the second supercapacitor through the third DC / DC converter.

[0016] The DC bus is connected to the AC power grid via a DC / AC converter.

[0017] The system is connected to the AC large power grid via a grid-connected controller.

[0018] The capacitance state sensing unit senses the state of charge value of the first super capacitor as S oc1 , sensing the state of charge value of the second supercapacitor is S oc2 ;

[0019] The power value output by the target photovoltaic energy panel through the first DC / DC converter is P1, the power value output by the photovoltaic energy panel to the left of the target photovoltaic energy panel through the second DC / DC converter is P2; the power value output by the photovoltaic energy panel to the right of the target photovoltaic energy panel through the third DC / DC converter is P3;

[0020] Then the battery unit is charged or discharged based on the charge state sensed by the capacitor state sensing unit, specifically including:

[0021] Calculate the power command value PB:

[0022]

[0023] Among them, S oc1,u , S oc1,l S oc1 The upper and lower limits of S oc2,u , S oc2,l Soc2 The upper and lower limits of

[0024] When PB is positive, the battery cell is discharged; otherwise, the battery cell is charged.

[0025] In a second aspect of the present invention, a photovoltaic power processing method is provided, the method is applied to the photovoltaic power processing system described in the first aspect, the photovoltaic power processing system includes photovoltaic energy combination units; each photovoltaic energy combination unit includes a target photovoltaic energy panel and a left adjacent photovoltaic energy panel and a right adjacent photovoltaic energy panel of the target photovoltaic energy panel;

[0026] The method comprises the following steps:

[0027] S100: Acquire a plurality of first historical power values ​​output by a target photovoltaic energy panel through a first DC / DC converter, a plurality of second historical power values ​​output by a photovoltaic energy panel adjacent to the left of the target photovoltaic energy panel through a second DC / DC converter, and a plurality of third historical power values ​​output by a photovoltaic energy panel adjacent to the right of the target photovoltaic energy panel through a third DC / DC converter;

[0028] S200: outputting a power prediction result of a photovoltaic energy panel adjacent to the target photovoltaic energy panel in a first preset time period through a static physical power prediction model based on numerical weather forecast data, topographic data, and a plurality of second historical power values;

[0029] S300: Outputting a power prediction result for a second preset time period based on a plurality of third historical power values ​​through a time series prediction model;

[0030] S400: Outputting a power prediction result for a third preset time period based on a plurality of first historical power values ​​through an artificial intelligence prediction model;

[0031] S500: When the power prediction result of the third preset time period includes the power prediction result of the first preset time period, or the power prediction result of the third preset time period includes the power prediction result of the second preset time period, triple the power prediction result of the third preset time period as the output power value of the photovoltaic energy combination unit in the future target preset time period;

[0032] S600: When the output power value matches the current output power of the AC large grid to be connected to the grid, the grid connection controller issues a grid connection instruction so that the photovoltaic energy combination unit is connected to the AC large grid to be connected to the grid;

[0033] The second preset time period is smaller than the first preset time period and the third preset time period; the third preset time period includes the future target preset time period and the first preset time period.

[0034] The photovoltaic power processing system also includes a capacitor unit, a battery unit and a capacitor state sensing unit;

[0035] The target photovoltaic energy panel is connected to the DC bus through a first DC / DC converter, the left neighboring photovoltaic energy panel of the target photovoltaic energy panel is connected to a second DC / DC converter, and the right neighboring photovoltaic energy panel of the target photovoltaic energy panel is connected to a third DC / DC converter;

[0036] The battery unit is connected to the target photovoltaic energy panel through a first DC / DC converter corresponding to the target photovoltaic energy panel, and is connected to the capacitor unit;

[0037] The left neighboring photovoltaic energy panel and the right neighboring photovoltaic energy panel of the target photovoltaic energy panel are connected to the battery unit through the capacitor unit;

[0038] The capacitor state sensing unit is used to sense the charge state of the capacitor unit;

[0039] The battery unit is charged or discharged based on the charge state sensed by the capacitance state sensing unit.

[0040] In step S600, when the difference between the sum of the output power value and the current output power of the AC large grid to be connected to the grid and the required power of the AC large grid to be connected to the grid is within a preset range, the output power matches the current output power of the AC large grid to be connected to the grid.

[0041] The step S500 also includes: when the power prediction result of the third preset time period includes neither the power prediction result of the first preset time period nor the power prediction result of the second preset time period, the artificial intelligence prediction model is retrained based on the power prediction result of the first preset time period and the power prediction result of the second preset time period until the training is completed, to obtain the retrained artificial intelligence prediction model, and return to step S400.

[0042] The length of the first preset time period is L1, the length of the second preset time period is L2, the length of the third preset time period is L3, and the length of L2 <L1,L2<L3;L3> L1; the third preset time period includes the first preset time period;

[0043] In step S500, the artificial intelligence prediction model is retrained based on the power prediction result of the first preset time period and the power prediction result of the second preset time period until the training is completed, and the conditions for completing the training include:

[0044] The training time is not less than the preset number Tnum.

[0045] The present application can realize intelligent power distribution, prediction and scheduling of a photovoltaic microgrid containing DC and AC hybrid loads. At the same time, the battery unit is charged or discharged based on the charge state sensed by the capacitor state sensing unit and the real-time output power or predicted output power of the target photovoltaic energy panel and the left adjacent photovoltaic energy panel and the right adjacent photovoltaic energy panel included in each photovoltaic energy combination unit, thereby enhancing the robustness of the microgrid power prediction, effectively improving the voltage output of the photovoltaic power source connected to the grid, and enhancing the reliability of the distribution network system.

[0046] Further advantages of the present invention will be further reflected in detail in the specific embodiments section in conjunction with the drawings of the specification. BRIEF DESCRIPTION OF THE DRAWINGS

[0047] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0048] Figure 1 This is a schematic diagram of the architecture of a photovoltaic power processing system according to an embodiment of the present invention.

[0049] Figure 2 yes Figure 1 Schematic diagram of some control principles of the photovoltaic power processing system

[0050] Figure 3 This is a schematic diagram of the main flow of a photovoltaic power processing method according to an embodiment of the present invention.

[0051] Figure 4 yes Figure 3 Schematic diagram of the training branches of the artificial intelligence model involved in the method DETAILED DESCRIPTION

[0052] Here, exemplary embodiments will be described in detail, examples of which are shown in the accompanying drawings. When the following description refers to the drawings, unless otherwise indicated, the same numbers in different drawings represent the same or similar elements. The embodiments described in the following exemplary embodiments do not represent all embodiments that are the same as the present application. Instead, they are only examples of devices and methods that are the same as some aspects of the present application as detailed in the attached claims.

[0053] In the embodiments of the present application, the term "module" or "unit" refers to a computer program or a part of a computer program with a predetermined function, and works together with other related parts to achieve a predetermined goal, and can be implemented in whole or in part by using software, hardware (such as processing circuits or memories) or a combination thereof. Similarly, a processor (or multiple processors or memories) can be used to implement one or more modules or units. In addition, each module or unit can be part of an overall module or unit that includes the function of the module or unit.

[0054] The block diagrams shown in the accompanying drawings are merely functional entities and do not necessarily correspond to physically independent entities. That is, these functional entities may be implemented in software form, or in one or more hardware modules or integrated circuits, or in different networks and / or processor devices and / or microcontroller devices.

[0055] Figure 1 It is a schematic diagram of the architecture of a photovoltaic power processing system according to an embodiment of the present invention.

[0056] exist Figure 1 In the system, the system includes a photovoltaic energy combination unit, a plurality of DC / DC converters, a capacitor unit, a battery unit, and a capacitor state sensing unit;

[0057] Each of the photovoltaic energy combination units includes a target photovoltaic energy panel and a left adjacent photovoltaic energy panel and a right adjacent photovoltaic energy panel of the target photovoltaic energy panel;

[0058] The target photovoltaic energy panel is connected to the DC bus through a first DC / DC converter, the left neighboring photovoltaic energy panel of the target photovoltaic energy panel is connected to a second DC / DC converter, and the right neighboring photovoltaic energy panel of the target photovoltaic energy panel is connected to a third DC / DC converter;

[0059] The battery unit is connected to the target photovoltaic energy panel through a first DC / DC converter corresponding to the target photovoltaic energy panel, and is connected to the capacitor unit;

[0060] The left neighboring photovoltaic energy panel and the right neighboring photovoltaic energy panel of the target photovoltaic energy panel are connected to the battery unit through the capacitor unit;

[0061] The capacitor state sensing unit is used to sense the charge state of the capacitor unit;

[0062] The battery unit is charged or discharged based on the charge state sensed by the capacitance state sensing unit.

[0063] Photovoltaic energy panels are an important part of solar photovoltaic power generation systems, also known as photovoltaic arrays.

[0064] A typical solar photovoltaic power generation system includes a photovoltaic array composed of photovoltaic modules, a photovoltaic junction box, a grid-connected inverter, an energy storage device and its charging control device (limited to systems with energy storage devices), an electric energy meter and instruments that display electric energy-related parameters.

[0065] In the independent island operation mode, the solar photovoltaic power generation system converts solar energy into electrical energy through solar panels and stores it in batteries, and then controls the inverter to convert the DC power in the battery into AC power for use by electrical equipment; in the grid-connected mode, after converting solar energy into electrical energy, the electrical energy is sent to the grid through the grid-connected inverter, realizing grid-connected operation with the large AC power grid.

[0066] In the prior art, the output power of a single photovoltaic array (the photovoltaic energy panel in this application) is usually predicted and adjusted, and only a single scheduling model and prediction method are used. However, due to the instability and intermittency of light, the photovoltaic output is volatile and uncontrollable. Therefore, the robustness of a single target control and a single model prediction method is poor.

[0067] To this end, Figure 1 In the embodiment, real-time control and predictive control are performed with the photovoltaic energy combination unit as the target.

[0068] Specifically, each of the photovoltaic energy combination units includes a target photovoltaic energy panel and a left-adjacent photovoltaic energy panel and a right-adjacent photovoltaic energy panel of the target photovoltaic energy panel.

[0069] The capacitor unit includes a first super capacitor and a second super capacitor;

[0070] The left neighboring photovoltaic energy panel is connected to the first supercapacitor via the second DC / DC converter;

[0071] The right-adjacent photovoltaic energy panel is connected to the second supercapacitor through the third DC / DC converter.

[0072] The target photovoltaic energy panel is connected to the DC bus through a first DC / DC converter, the left neighboring photovoltaic energy panel of the target photovoltaic energy panel is connected to a second DC / DC converter, and the right neighboring photovoltaic energy panel of the target photovoltaic energy panel is connected to a third DC / DC converter;

[0073] Combination Figure 2The control principle diagram, in the real-time control mode, the power value of the target photovoltaic energy panel output through the first DC / DC converter is P1, the power value of the photovoltaic energy panel to the left of the target photovoltaic energy panel output through the second DC / DC converter is P2; the power value of the photovoltaic energy panel to the right of the target photovoltaic energy panel output through the third DC / DC converter is P3;

[0074] The battery unit is charged or discharged based on the charge state sensed by the capacitor state sensing unit and the real-time output power or predicted output power of the target photovoltaic energy panel and the left and right adjacent photovoltaic energy panels of the target photovoltaic energy panel included in each photovoltaic energy combination unit.

[0075] In an embodiment of the present application, the battery unit is preferably a chargeable / dischargeable lithium battery, such as a lithium-ion battery pack.

[0076] Generally speaking, supercapacitors have the advantages of high power density, long cycle life, high efficiency, and high current charging and discharging without affecting life, but low energy density; while lithium-ion batteries have high energy density, so in the example of this application, lithium-ion batteries with high energy density and supercapacitors with high power density are combined into a hybrid energy storage system, which can make the energy storage system have both good power characteristics and energy characteristics. However, this raises the problem of controlling the charging and discharging state of lithium-ion battery packs.

[0077] Existing technologies usually adjust the charging and discharging status of lithium-ion battery packs based on the change instructions of the total output power of the photovoltaic microgrid. This method has limited control flexibility and is also unstable based on the dependent total output power value. The control strategy is not robust enough, and in extreme cases it leads to excessive charging and discharging rates, which seriously affects the battery life.

[0078] To this end, supercapacitors with better power characteristics are used to respond to the overall short-term high-power instructions of the energy storage system, while batteries that are easier to reach large capacity are used to meet the energy demand of the energy storage system over a longer time scale. Supercapacitors respond to peak power and can avoid high-rate charge and discharge and deep discharge of lithium-ion batteries in a short period of time. In certain appropriate power instruction fluctuations, the capacity of supercapacitors is fully utilized to minimize the number of charge and discharge cycles of lithium-ion batteries, thereby extending the service life of lithium-ion batteries to a certain extent.

[0079] Based on this, when based on real-time power control, the battery unit is charged or discharged based on the state of charge sensed by the capacitor state sensing unit, specifically including:

[0080] Calculate the power command value PB:

[0081]

[0082] Among them, S oc1,u , S oc1,l S oc1 The upper and lower limits of S oc2,u , S oc2,l S oc2 The upper and lower limits of

[0083] When PB is positive, the battery cell is discharged; otherwise, the battery cell is charged.

[0084]

[0085] The capacitance state sensing unit senses the state of charge value of the first supercapacitor as S oc1 , sensing the state of charge value of the second supercapacitor is S oc2 ; state of charge (S OC ) value reflects the cumulative effect of current in the supercapacitor, and its change is slower than that of current and power.

[0086] Based on the power distribution method controlled by the above PB value, the power command of the battery changes smoothly with the SOC of the supercapacitor, so that the capacity of the supercapacitor can be used to smooth the fluctuation of the battery power;

[0087] In addition to real-time control, a further preferred embodiment of the present application also includes a predictive control mode based on power combination prediction.

[0088] Specifically, the power combination prediction includes static physical power prediction model, time series prediction model and artificial intelligence prediction model;

[0089] Among them, the artificial intelligence prediction model predicts the output power of the target photovoltaic energy panel, and the static physical power prediction model and the time series prediction model predict the output power of the left neighboring photovoltaic energy panel and the right neighboring photovoltaic energy panel of the target photovoltaic energy panel respectively.

[0090] The static physical power prediction model builds a power prediction model based on geographic information, meteorological data and numerical weather forecast results. In photovoltaic power prediction, detailed geographic information and related meteorological information of photovoltaic power stations are collected, and then the prediction results are calculated through physical equations related to solar radiation transmission and photovoltaic device operation.

[0091] The time series prediction model predicts future photovoltaic,output based on historical power data and is mainly used for short-term and,ultra-short-term power prediction.

[0092] The artificial intelligence prediction model learns the mapping relationship between input data and output new energy power prediction results through a large amount of historical training data, including neural networks, Markov processes, support vector machines, extreme learning machines, fuzzy logic, etc. The artificial intelligence method does not rely on any physical model, can automatically complete feature extraction of input data, and has good generalization performance when there are sufficient training samples. Therefore, it is widely used in ultra-short-term, short-term, medium-term and long-term predictions.

[0093] For the specific model principles of static physical power prediction, time series prediction, and artificial intelligence prediction, please refer to the prior art, and this application will not elaborate on this. For example, the following prior art has detailed records of the principles of the above-mentioned different prediction models themselves, which are introduced as part of the specific embodiments of this application:

[0094] [1]ALMONACID F, et al. A methodology based on dynamic artificial neural network for short-term forecasting of the power output of aPV generator[J]. Energy Conversion and Management, 2014(85): 389-398.

[0095] [2] Xue Yusheng, et al. Review on short-term and ultra-short-term wind power forecasting [J]. Automation of Electric Power Systems, 2015, 39(6): 141-151.

[0096] [3]MENG Z, et al. Nonparametric multivariate probability density forecast in smart grids with deep learning [J / OL]. IEEE Transactions on PowerSystems, 2022.

[0097] On this basis, the predictive control mode based on power combination prediction is specifically implemented as follows:

[0098] S1: Acquire a plurality of first historical power values ​​output by a target photovoltaic energy panel through a first DC / DC converter, a plurality of second historical power values ​​output by a photovoltaic energy panel adjacent to the left of the target photovoltaic energy panel through a second DC / DC converter, and a plurality of third historical power values ​​output by a photovoltaic energy panel adjacent to the right of the target photovoltaic energy panel through a third DC / DC converter;

[0099] S2: Outputting the power prediction result (P20) of the left neighboring photovoltaic energy panel of the target photovoltaic energy panel in the first preset time period based on numerical weather forecast data, topographic data, and multiple second historical power values ​​through a static physical power prediction model;

[0100] S3: Outputting a power prediction result (P30) for a second preset time period based on a plurality of third historical power values ​​through a time series prediction model;

[0101] S4: Outputting a power prediction result (P10) for a third preset time period based on the plurality of first historical power values ​​through an artificial intelligence prediction model;

[0102] At this time, continue to calculate the power command value PB:

[0103]

[0104] When PB is positive, the battery cell is discharged; otherwise, the battery cell is charged.

[0105] In practical applications, a plurality of DC loads are directly connected to the DC bus; the DC bus is connected to a large AC power grid via a DC / AC converter.

[0106] Preferably, the system is connected to the AC large grid via a grid-connected controller, and the grid-connected controller can realize isolation control of the photovoltaic microgrid system and the AC large grid to be connected to the grid.

[0107] Next, the method embodiment of the technical solution of the present application is introduced. It can be understood that after the above detailed description of the embodiment in the form of system flow, the corresponding embodiment in the form of method flow does not need to be repeated again, but those skilled in the art can directly and unambiguously determine that the embodiment in the form of method flow also corresponds to all the contents of the embodiment in the form of system functional architecture. The same is true for the embodiment in the form of computer program code pseudo language.

[0108] Figure 3 It is a schematic diagram of the main flow of a photovoltaic power processing method according to an embodiment of the present invention.

[0109] Understandably, Figure 3 The embodiment of the present invention focuses on the combined power prediction problem in photovoltaic power processing. Figure 1 The existing contents of the embodiments will not be repeated. Figure 1 and Figure 3 Different embodiments belonging to the same inventive concept.

[0110] Specifically, Figure 3The method is applied to the photovoltaic power processing system described in the first aspect, wherein the photovoltaic power processing system comprises photovoltaic energy combination units; each photovoltaic energy combination unit comprises a target photovoltaic energy panel and a left adjacent photovoltaic energy panel and a right adjacent photovoltaic energy panel of the target photovoltaic energy panel;

[0111] The method comprises the following steps:

[0112] S100: Acquire a plurality of first historical power values ​​output by a target photovoltaic energy panel through a first DC / DC converter, a plurality of second historical power values ​​output by a photovoltaic energy panel adjacent to the left of the target photovoltaic energy panel through a second DC / DC converter, and a plurality of third historical power values ​​output by a photovoltaic energy panel adjacent to the right of the target photovoltaic energy panel through a third DC / DC converter;

[0113] S200: outputting a power prediction result of a photovoltaic energy panel adjacent to the target photovoltaic energy panel in a first preset time period through a static physical power prediction model based on numerical weather forecast data, topographic data, and a plurality of second historical power values;

[0114] S300: Outputting a power prediction result for a second preset time period based on a plurality of third historical power values ​​through a time series prediction model;

[0115] S400: Outputting a power prediction result for a third preset time period based on a plurality of first historical power values ​​through an artificial intelligence prediction model;

[0116] S500: When the power prediction result of the third preset time period includes the power prediction result of the first preset time period, or the power prediction result of the third preset time period includes the power prediction result of the second preset time period, triple the power prediction result of the third preset time period as the output power value of the photovoltaic energy combination unit in the future target preset time period;

[0117] S600: When the output power value matches the current output power of the AC large grid to be connected to the grid, the grid connection controller issues a grid connection instruction so that the photovoltaic energy combination unit is connected to the AC large grid to be connected to the grid;

[0118] The second preset time period is smaller than the first preset time period and the third preset time period; the third preset time period includes the future target preset time period and the first preset time period.

[0119] The photovoltaic power processing system also includes a capacitor unit, a battery unit and a capacitor state sensing unit;

[0120] The target photovoltaic energy panel is connected to the DC bus through a first DC / DC converter, the left neighboring photovoltaic energy panel of the target photovoltaic energy panel is connected to a second DC / DC converter, and the right neighboring photovoltaic energy panel of the target photovoltaic energy panel is connected to a third DC / DC converter;

[0121] The battery unit is connected to the target photovoltaic energy panel through a first DC / DC converter corresponding to the target photovoltaic energy panel, and is connected to the capacitor unit;

[0122] The left neighboring photovoltaic energy panel and the right neighboring photovoltaic energy panel of the target photovoltaic energy panel are connected to the battery unit through the capacitor unit;

[0123] The capacitor state sensing unit is used to sense the charge state of the capacitor unit;

[0124] The battery unit is charged or discharged based on the charge state sensed by the capacitance state sensing unit.

[0125] In step S600, when the difference between the sum of the output power value and the current output power of the AC large grid to be connected to the grid and the required power of the AC large grid to be connected to the grid is within a preset range, the output power matches the current output power of the AC large grid to be connected to the grid.

[0126] Figure 4 Further showing Figure 3 Schematic diagram of the training branches of the artificial intelligence model involved in the method.

[0127] Right now Figure 4 It is shown that step S500 also includes: when the power prediction result of the third preset time period includes neither the power prediction result of the first preset time period nor the power prediction result of the second preset time period, the artificial intelligence prediction model is retrained based on the power prediction result of the first preset time period and the power prediction result of the second preset time period until the training is completed, and the retrained artificial intelligence prediction model is obtained, and then return to step S400.

[0128] The length of the first preset time period is L1, the length of the second preset time period is L2, the length of the third preset time period is L3, and the length of L2 <L1,L2<L3;L3> L1; the third preset time period includes the first preset time period;

[0129] In step S500, the artificial intelligence prediction model is retrained based on the power prediction result of the first preset time period and the power prediction result of the second preset time period until the training is completed, and the conditions for completing the training include:

[0130] The training time is not less than the preset number Tnum;

[0131] in represents rounding up, and Δt represents the average time spent on each training.

[0132] It can be seen that the calculation of the above preset number of times comprehensively considers the prediction scales of different combined prediction models themselves, and can better balance the different prediction scales that affect the training model.

[0133] As a general technical solution, combined with Figure 1 and Figure 3 A more complete embodiment of the technical solution of the present application includes:

[0134] The photovoltaic power processing system has a real-time control mode and a combined predictive control mode;

[0135] In the real-time control mode, the capacitor state sensing unit senses the state of charge value of the first super capacitor as S oc1 , sensing the state of charge value of the second supercapacitor is S oc2 ;

[0136] The power value output by the target photovoltaic energy panel through the first DC / DC converter is P1, the power value output by the photovoltaic energy panel to the left of the target photovoltaic energy panel through the second DC / DC converter is P2; the power value output by the photovoltaic energy panel to the right of the target photovoltaic energy panel through the third DC / DC converter is P3;

[0137] Then the battery unit is charged or discharged based on the charge state sensed by the capacitor state sensing unit, specifically including:

[0138] Calculate the power command value PB:

[0139]

[0140] In the combined predictive control mode,

[0141] First, a plurality of first historical power values ​​output by a target photovoltaic energy panel through a first DC / DC converter, a plurality of second historical power values ​​output by a photovoltaic energy panel adjacent to the left of the target photovoltaic energy panel through a second DC / DC converter, and a plurality of third historical power values ​​output by a photovoltaic energy panel adjacent to the right of the target photovoltaic energy panel through a third DC / DC converter are obtained;

[0142] Secondly, the static physical power prediction model is used to output the power prediction result (P20) of the left neighboring photovoltaic energy panel of the target photovoltaic energy panel in the first preset time period based on numerical weather forecast data, topographic data, and multiple second historical power values; the time series prediction model is used to output the power prediction result (P30) of the second preset time period based on multiple third historical power values; the artificial intelligence prediction model is used to output the power prediction result (P10) of the third preset time period based on multiple first historical power values;

[0143] At this time, calculate the power command value PB:

[0144]

[0145] When PB is positive, the battery cell is discharged; otherwise, the battery cell is charged.

[0146] Actual scheduling data tests have proved that by configuring each of the photovoltaic energy combination units to include a target photovoltaic energy panel and the left adjacent photovoltaic energy panel and the right adjacent photovoltaic energy panel of the target photovoltaic energy panel and counting the power value of the target photovoltaic energy panel output through the first DC / DC converter as P1, the power value of the left adjacent photovoltaic energy panel of the target photovoltaic energy panel output through the second DC / DC converter as P2; the power value of the right adjacent photovoltaic energy panel of the target photovoltaic energy panel output through the third DC / DC converter as P3, the power command value PB is calculated in real-time mode, and the battery unit is charged or discharged based on the charge state sensed by the capacitor state sensing unit and the real-time output power or predicted output power of the target photovoltaic energy panel and the left adjacent photovoltaic energy panel and the right adjacent photovoltaic energy panel of each photovoltaic energy combination unit, thereby avoiding high-rate charging and discharging and deep discharge of lithium-ion batteries in a short period of time. The number of charge and discharge cycles of the lithium-ion battery is reduced as much as possible, and the service life of the lithium-ion battery is extended to a certain extent; and when the output power value matches the current output power of the large AC power grid to be connected to the grid, the grid-connected controller issues a grid-connected instruction, so that the photovoltaic energy combination unit is connected to the large AC power grid to be connected to the grid, which can realize the intelligent power distribution, prediction and scheduling of the photovoltaic microgrid containing DC and AC mixed loads, enhance the robustness of the microgrid power prediction, effectively improve the voltage output of the photovoltaic power grid, and enhance the reliability of the distribution network system.

[0147] The present invention provides multiple embodiments, each of which can constitute an independent technical solution and may contribute to the prior art and solve corresponding technical problems. However, it should be pointed out that different embodiments can be combined with each other without violating logic; at the same time, each embodiment can solve at least one technical problem, but it is not required that each individual embodiment solves multiple or all technical problems.

[0148] When the embodiments of the present application are applied to specific products or technologies, user permission or consent is required, and the collection, use and processing of relevant data must comply with relevant laws, regulations and standards of relevant countries and regions.

[0149] The foregoing has shown and described the method embodiments and system of the present invention, but it is understandable to those skilled in the art that various changes, modifications, substitutions and variations may be made to these embodiments without departing from the principles and spirit of the present invention, and the scope of the present invention is defined by the appended claims and their equivalents.

Claims

1. A photovoltaic power processing system, the system comprising a photovoltaic energy combination unit, a plurality of DC / DC converters, a capacitor unit, a battery unit and a capacitor state sensing unit; Features: Each of the photovoltaic energy combination units includes a target photovoltaic energy panel and a left adjacent photovoltaic energy panel and a right adjacent photovoltaic energy panel of the target photovoltaic energy panel; The target photovoltaic energy panel is connected to the DC bus through a first DC / DC converter, the left neighboring photovoltaic energy panel of the target photovoltaic energy panel is connected to a second DC / DC converter, and the right neighboring photovoltaic energy panel of the target photovoltaic energy panel is connected to a third DC / DC converter; The battery unit is connected to the target photovoltaic energy panel through a first DC / DC converter corresponding to the target photovoltaic energy panel, and is connected to the capacitor unit; The left neighboring photovoltaic energy panel and the right neighboring photovoltaic energy panel of the target photovoltaic energy panel are connected to the battery unit through the capacitor unit; The capacitor state sensing unit is used to sense the charge state of the capacitor unit; The battery unit is charged or discharged based on the charge state sensed by the capacitance state sensing unit; The capacitor unit includes a first super capacitor and a second super capacitor; The left neighboring photovoltaic energy panel is connected to the first supercapacitor via the second DC / DC converter; The right adjacent photovoltaic energy panel is connected to the second super capacitor via the third DC / DC converter; The capacitance state sensing unit senses the state of charge value of the first super capacitor as S oc1 , sensing the state of charge value of the second supercapacitor is S oc2 ; The power value output by the target photovoltaic energy panel through the first DC / DC converter is P1, the power value output by the photovoltaic energy panel to the left of the target photovoltaic energy panel through the second DC / DC converter is P2; the power value output by the photovoltaic energy panel to the right of the target photovoltaic energy panel through the third DC / DC converter is P3; Then the battery unit is charged or discharged based on the charge state sensed by the capacitor state sensing unit, specifically including: Calculate the power command value PB: Among them, S oc1,u , S oc1,l S oc1 The upper and lower limits of S oc2,u , S oc2,l S oc2 The upper and lower limits of When PB is positive, the battery cell is discharged; otherwise, the battery cell is charged.

2. A photovoltaic power processing system as claimed in claim 1, characterized in that: A plurality of DC loads are directly connected to the DC bus; The DC bus is connected to the AC power grid via a DC / AC converter.

3. A photovoltaic power processing system as claimed in claim 1, characterized in that: The system is connected to the AC large power grid via a grid-connected controller.

4. A photovoltaic power processing method, the method being applied to the photovoltaic power processing system according to any one of claims 1 to 3, It is characterized in that The method comprises the following steps: S100: Acquire a plurality of first historical power values ​​output by a target photovoltaic energy panel through a first DC / DC converter, a plurality of second historical power values ​​output by a photovoltaic energy panel adjacent to the left of the target photovoltaic energy panel through a second DC / DC converter, and a plurality of third historical power values ​​output by a photovoltaic energy panel adjacent to the right of the target photovoltaic energy panel through a third DC / DC converter; S200: outputting a power prediction result of a photovoltaic energy panel adjacent to the target photovoltaic energy panel in a first preset time period through a static physical power prediction model based on numerical weather forecast data, topographic data, and a plurality of second historical power values; S300: Outputting a power prediction result for a second preset time period based on a plurality of third historical power values ​​through a time series prediction model; S400: Outputting a power prediction result for a third preset time period based on a plurality of first historical power values ​​through an artificial intelligence prediction model; S5 00: when the power prediction result of the third preset time period includes the power prediction result of the first preset time period, or the power prediction result of the third preset time period includes the power prediction result of the second preset time period, three times the power prediction result of the third preset time period is used as the output power value of the photovoltaic energy combination unit in the future target preset time period; S600: When the output power value matches the current output power of the AC large grid to be connected to the grid, the grid connection controller issues a grid connection instruction so that the photovoltaic energy combination unit is connected to the AC large grid to be connected to the grid; The second preset time period is smaller than the first preset time period and the third preset time period; the third preset time period includes the future target preset time period and the first preset time period.

5. A photovoltaic power processing method as claimed in claim 4, characterized in that: In step S600, when the difference between the sum of the output power value and the current output power of the AC large grid to be connected to the grid and the required power of the AC large grid to be connected to the grid is within a preset range, the output power matches the current output power of the AC large grid to be connected to the grid.

6. A photovoltaic power processing method as claimed in claim 5, characterized in that: The step S500 also includes: when the power prediction result of the third preset time period includes neither the power prediction result of the first preset time period nor the power prediction result of the second preset time period, the artificial intelligence prediction model is retrained based on the power prediction result of the first preset time period and the power prediction result of the second preset time period until the training is completed, to obtain the retrained artificial intelligence prediction model, and return to step S400.

7. A photovoltaic power processing method as claimed in claim 6, characterized in that: The length of the first preset time period is L1, the length of the second preset time period is L2, the length of the third preset time period is L3, and the length of L2 <L1,L2<L3;L3> L1; the third preset time period includes the first preset time period.

Images