Photovoltaic power supply device and method

Abstract

The present invention provides a photovoltaic power supply device and method, comprising: a photovoltaic cell, a battery charging management module, a battery pack, a boost circuit module, a main control board, an information acquisition circuit module, a buck (buck) direct current (DC) module, and an inverter module; the battery charging management module controls the photovoltaic cell to charge the battery pack using MPPT output; the information acquisition circuit module collects the load voltage of the buck direct current (DC) module and the inverter module; and the main control board determines a square wave pulse width ratio and a sine wave pulse width ratio based on the load voltage and target output voltage, thereby adjusting the actual output voltage of the buck direct current (DC) module and the inverter module based on the square wave pulse width ratio and the sine wave pulse width ratio. The photovoltaic power supply device provided by the present invention has AC and DC voltage outputs and is adaptable to multiple voltage levels. It uses the LS-SVM method, utilizing the high-precision and high-stability regression method to automatically track the actual voltage requirements of the device, meeting the requirements for high-precision voltage matching.

Description

Technical Field

[0001] The present invention relates to the field of power electronics technology, and in particular to a photovoltaic power supply device and method. Background Art

[0002] For smart parks, which have been comprehensively upgraded through the integration of next-generation information technologies such as the Internet of Things, cloud computing, and big data, scientific and rational overall planning is particularly important. Currently, smart parks rely primarily on dedicated municipal power lines for power supply. However, given the diverse functional structures of the park, photovoltaic equipment is required to generate power independently to meet the park's individual needs. Because different devices have varying power supply voltage requirements, the need for power supply equipment to adapt to different voltage levels is becoming increasingly prominent. Current research on photovoltaic equipment primarily addresses power conversion efficiency, specifically constant-voltage output systems, with limited research examining system adaptation to multiple voltage levels.

[0003] The main voltage required by campus equipment ranges from 0 to 24V. Most power supplies on the market are specific, and a power supply module is customized between different devices. For adaptive power supplies, the least squares method is mainly used to adapt the power supply of various equipment in the campus, and the error is about 0.1V.

[0004] Due to the large power supply error of the equipment, there are fluctuations in electrical energy, which will cause overvoltage and undervoltage in subsequent circuits, and will cause damage to the equipment in the long run. Summary of the Invention

[0005] In view of the problems existing in the prior art, embodiments of the present invention provide a photovoltaic power supply device and method.

[0006] In a first aspect, the present invention provides a photovoltaic power supply device, comprising: a photovoltaic cell, a battery charging management module, a battery pack, a BOOST circuit module, a main control board, an information acquisition circuit module, a BUCK DC module, and an inverter module; the battery charging management module is used to control the photovoltaic cell to charge the battery pack using a maximum power point tracking method; the input end of the BOOST circuit module is connected to the output end of the battery pack; the BUCK DC module and the inverter module are respectively connected to the output end of the BOOST circuit module; the information acquisition circuit module is used to collect the load voltages of the BUCK DC module and the inverter module, and feed the load voltages back to the main control board; the main control board re-determines the square wave pulse width ratio of the BUCK DC module and the sine wave pulse width ratio of the inverter module based on the load voltage and the target output voltages of the BUCK DC module and the inverter module, and feeds back the information to the battery charging management module; the battery charging management module adjusts the actual output voltages of the BUCK DC module and the inverter module based on the square wave pulse width ratio and the sine wave pulse width ratio.

[0007] In one embodiment, the main control board is used to run a trained least squares support vector machine model; the least squares support vector machine model is used to perform a linear regression operation on the load voltage and the target output voltage to obtain fitting parameters; the main control board re-determines the square wave pulse width ratio of the BUCK DC module and the sine wave pulse width ratio of the inverter module based on the fitting parameters. In one embodiment, the inverter module includes an IR2104 drive circuit, an LC low-pass filter circuit, and a switching circuit; the main control board outputs a bipolar sine wave to the control end of the IR2104 drive circuit through a table lookup method, and controls the cross-conduction of the bipolar sine wave to control the switching frequency of the switching circuit; the LC low-pass filter circuit is used to receive the DC power input by the BUCK DC module and output AC power; the switching frequency of the switching circuit is used to adjust the output amplitude of the AC power.

[0008] In one embodiment, the control chip of the BOOST circuit module is a TPS61175 chip.

[0009] In one embodiment, the BUCK DC module includes an H-bridge driving circuit.

[0010] In one embodiment, there are multiple BUCK DC modules and / or multiple inverter modules.

[0011] In one embodiment, the photovoltaic power supply device further includes a host computer, which is communicatively connected to the main control board; the host computer is used to receive and store the historical load voltage of each of the BUCK DC modules and each of the inverter modules collected by the information acquisition circuit module, and construct a training set and a test set; and the least squares support vector machine model is pre-trained using the training set and the test set.

[0012] In a second aspect, the present invention provides a photovoltaic power supply method, comprising the following steps:

[0013] Step S1, determining an initial square wave pulse width ratio of the BUCK DC module and an initial sine wave pulse width ratio of the inverter module, so that the battery charging management module controls the BUCK DC module to output a corresponding initial DC power, and controls the inverter module to output a corresponding initial AC power;

[0014] Step S2, receiving the load voltages output by the BUCK DC module and the inverter module collected by the information collection circuit module; the load voltages include the DC load voltage output by the BUCK DC module and the AC load voltage output by the inverter module;

[0015] Step S3: Sending the load voltage to the main control board, so that the main control board can redetermine a new square wave pulse width ratio of the buck DC module based on the DC load voltage and the target output voltage of the buck DC module; and redetermine a new sinusoidal wave pulse width ratio of the inverter module based on the AC load voltage and the target output voltage of the inverter module;

[0016] Step S4, feeding back the new square wave pulse width ratio and the new sine wave pulse width ratio to the battery charging management module, so that the battery charging management module adjusts the actual output voltages of the BUCK DC module and the inverter module respectively;

[0017] Step S5, iteratively execute steps S1 to S4.

[0018] In one embodiment, the main control board re-determines a new square wave pulse width ratio of the buck DC module based on the DC load voltage and the target output voltage of the buck DC module; and re-determines a new sinusoidal wave pulse width ratio of the inverter module based on the AC load voltage and the target output voltage of the inverter module, including:

[0019] Based on a pre-trained least squares support vector machine model, with the goal of controlling the photovoltaic cell to charge the battery pack in a maximum power point tracking manner, linear regression operations are performed on the DC load voltage and the AC load voltage with the corresponding target output voltage to obtain corresponding fitting parameters;

[0020] The square wave pulse width ratio of the BUCK DC module and the sine wave pulse width ratio of the inverter module are re-determined according to the fitting parameters.

[0021] In one embodiment, the number of data features of voltage levels is increased to construct training sets and test sets;

[0022] Based on the k-fold cross validation method and while reducing the high-order effects, the least squares support vector machine model is pre-trained using the training set and the test set.

[0023] In a third aspect, the present invention provides an electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein when the processor executes the program, the steps of any of the photovoltaic power supply devices described above are implemented.

[0024] In a fourth aspect, the present invention further provides a non-transitory computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the steps of any of the photovoltaic power supply devices described above.

[0025] The photovoltaic power supply device and method provided by the present invention have AC and DC voltage outputs and can adapt to multiple voltage levels. They adopt the LS-SVM method and automatically track the actual voltage requirements of the equipment under the high-precision and high-stability calculation of the regression method to meet the requirements of high-precision voltage matching. The power supply tracking error can be controlled at the 0.01V level. BRIEF DESCRIPTION OF THE DRAWINGS

[0026] In order to more clearly illustrate the technical solutions in the present invention or the prior art, a brief introduction is given below to the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0027] Figure 1 It is a structural schematic diagram of the photovoltaic power supply device provided by the present invention;

[0028] Figure 2 This is a schematic diagram of the campus power adapter provided by the present invention;

[0029] Figure 3 It is a schematic diagram of the flow of voltage regression control provided by the present invention;

[0030] Figure 4 This is a circuit diagram of a battery charging management module provided by the present invention;

[0031] Figure 5 This is a circuit diagram of a BOOST circuit module provided by the present invention;

[0032] Figure 6 This is a circuit diagram of a BUCK DC module provided by the present invention;

[0033] Figure 7 This is a schematic diagram of the process of loading the LV-SVM model on a software and hardware platform provided by the present invention;

[0034] Figure 8 This is a schematic flow chart of the photovoltaic power supply method provided by the present invention;

[0035] Figure 9 This is a diagram illustrating the relative usage of test sets and training sets in traditional model training;

[0036] Figure 10 Schematic diagram of the relative usage of the test set and the training set in model training using k-fold cross validation provided by the present invention;

[0037] Figure 11 This is a diagram of the training effect obtained by using different K values;

[0038] Figure 12 This is a diagram of the training and prediction results;

[0039] Figure 13 It is a schematic diagram of the prediction results of the test set;

[0040] Figure 14 It is a structural schematic diagram of the electronic device provided by the present invention. DETAILED DESCRIPTION

[0041] To make the objectives, technical solutions, and advantages of the present invention more clear, the technical solutions of the present invention will be clearly and completely described below in conjunction with the accompanying drawings. Obviously, the embodiments described are only some of the embodiments of the present invention, not all of them. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts shall fall within the scope of protection of the present invention.

[0042] It should be noted that in the description of the embodiments of the present invention, the terms "comprise," "include," or any other variations thereof are intended to encompass non-exclusive inclusion, such that a process, device, article, or apparatus comprising a series of elements includes not only those elements but also other elements not explicitly listed, or includes elements inherent to such process, device, article, or apparatus. Without further limitation, an element defined by the phrase "comprises a..." does not preclude the presence of additional identical elements in the process, device, article, or apparatus comprising the element. Terms such as "upper" and "lower" indicate positions or location relationships based on those shown in the accompanying drawings and are intended solely for ease of description and simplification of the present invention. They are not intended to indicate or imply that the method or element referred to must have a specific orientation, be constructed, or operate in a specific orientation, and are therefore not to be construed as limiting the present invention. Unless otherwise expressly specified or limited, the terms "mounted," "connected," and "connected" should be broadly construed, for example, to mean a fixed connection, a detachable connection, or an integral connection; a mechanical connection or an electrical connection; a direct connection or an indirect connection through an intermediate medium; or internal communication between two elements. For those skilled in the art, the specific meanings of the above terms in the present invention can be understood according to specific circumstances.

[0043] The following combination Figures 1-14 The photovoltaic power supply device and method provided by the embodiments of the present invention are described.

[0044] Figure 1 This is a schematic diagram of the structure of the photovoltaic power supply device provided by the present invention. Figure 1As shown, it mainly includes: photovoltaic cells, battery charging management module, battery pack, BOOST circuit module, main control board, information acquisition circuit module, BUCK DC module and inverter module;

[0045] The battery charging management module is used to control the photovoltaic cell to charge the battery pack in a maximum power point tracking manner;

[0046] The input end of the BOOST circuit module is connected to the output end of the battery pack;

[0047] The BUCK DC module and the inverter module are respectively connected to the output end of the BOOST circuit module;

[0048] The information acquisition circuit module is used to collect the load voltage of the BUCK DC module and the inverter module, and feed the load voltage back to the main control board;

[0049] The main control board re-determines the square wave pulse width ratio of the BUCK DC module and the sine wave pulse width ratio of the inverter module according to the load voltage and the target output voltages of the BUCK DC module and the inverter module, and feeds back the information to the battery charging management module;

[0050] The battery charging management module adjusts the actual output voltages of the BUCK DC module and the inverter module respectively according to the square wave pulse width ratio and the sine wave pulse width ratio.

[0051] Because the photovoltaic power supply provided by existing technologies, on the one hand, focuses on the maximum power tracking algorithm, the research does not take into account the real needs of the park, which is the adaptation requirements of voltage accuracy. In addition, most existing solutions are constant voltage output, which cannot meet the adaptation requirements of 0~24V multiple voltage levels of many low-power equipment in the park; on the other hand, most existing solutions are DC-DC (direct current-direct current) step-down systems, which have poor compatibility with AC systems.

[0052] To address the above shortcomings, the photovoltaic power supply device provided by the present invention has both AC and DC voltage outputs, and outputs maximum power by adopting the Maximum Power Point Tracking (MPPT) method through the battery charging management module while meeting the requirement of low power consumption.

[0053] MPPT (Medium Power Permit) is a technology that adjusts the operating state of electrical modules to maximize the power output of photovoltaic cells, effectively storing the DC power generated by the cells in a battery pack. The output power of a photovoltaic cell is related to the operating voltage of the controller where the MPPT is installed. Only when operating at the optimal voltage can its output power reach its maximum value.

[0054] Optionally, the main control board mainly includes a microcontroller unit (MCU), and the present invention uses low-power control processing for it, that is, the MCU enters a low-power mode when it is idle to meet the low-power consumption requirements of the photovoltaic system.

[0055] The photovoltaic power supply device provided by the present invention is mainly divided into the following parts: photovoltaic cells, a battery charging management module capable of performing MPPT, a battery pack, a BOOST circuit module, a BUCK DC module, an inverter module, a main control board and an information collection circuit module.

[0056] The main circuit includes: a boost circuit module and a buck DC module, based on pulse width modulation (PWM) to achieve DC-DC conversion, thereby converting the DC power output of the battery pack into different voltage outputs; and a boost circuit module and an inverter module, based on sinusoidal pulse width modulation (SPWM) to achieve DC-AC conversion, thereby converting the DC power output of the battery pack into the required alternating current output. Therefore, the photovoltaic power supply device provided by the present invention can meet the operating requirements of different low-power devices by constructing DC-DC and DC-AC systems.

[0057] It should be noted that the photovoltaic power supply device provided by the present invention also uses an information acquisition circuit module to collect the load voltage output by the buck DC module, and returns the collected load voltage to the target output voltage initially set by the buck DC module, thereby closed-loop adjusting the square wave pulse width ratio of the buck DC module to adjust its output voltage, so that it automatically tracks the power consumption of the equipment connected to the buck DC module to quickly match the voltage level required by the user.

[0058] Similarly, the photovoltaic power supply device provided by the present invention also collects the load voltage output by the inverter module through the information collection circuit module, and returns the collected load voltage to the target output voltage initially set by the inverter module, to close-loop adjust the sinusoidal wave pulse width ratio of the inverter module to adjust its output voltage, so that it automatically tracks the power consumption of the equipment connected to the inverter module to quickly match the voltage level required by the user.

[0059] Among them, the information acquisition circuit module mainly includes a feedback circuit.

[0060] In addition, the present invention uses the photovoltaic cell as the input end and controls the MPPT output through the battery charging management module to extend the service life of the photovoltaic cell.

[0061] The photovoltaic power supply device provided by the present invention has AC and DC voltage outputs and can adapt to multiple voltage levels. It adopts the LS-SVM method and automatically tracks the actual voltage requirements of the equipment under the high-precision and high-stability calculation of the regression method to meet the needs of high-precision voltage matching. The power supply tracking error can be controlled at the 0.01V level.

[0062] Based on the content of the above embodiment, as an optional embodiment, there are multiple BUCK DC modules and / or multiple inverter modules.

[0063] Figure 2 This is a schematic diagram of the campus power adapter provided by the present invention, such as Figure 2 As shown, the park's power transmission network consists of two types: one uses a dedicated 10kV municipal power line, which is converted to 220V for civilian use in the power distribution room. The park's main power supply is 220V, which is then converted to DC by an inverter to power equipment. The other type of power supply is a constant power source output by photovoltaic power supply devices, generally with a fixed voltage level. Specifically, the power supply for companies throughout the park primarily consists of 220V AC power provided by the park's power distribution room and DC power provided by the photovoltaic power supply devices.

[0064] The DC power output from the output end of the battery pack is controlled at 24V after passing through the BOOST circuit module. Figure 2 There are five devices (device 1 to device 5), including four DC devices whose operating voltages (equivalent to the target output voltage) are 3.3V, 5V, 12V and 24V respectively; and one AC device whose operating voltage is 12V.

[0065] Optionally, the output end of the battery pack is controlled at 24V after passing through a BOOST circuit module.

[0066] Four BUCK DC modules and one inverter module are connected to the battery pack to build corresponding DC-DC and DC-AC circuits. The main control board outputs corresponding square waves and sine waves through a lookup table method, so that the battery charging management module controls the battery pack to output the corresponding DC or AC power.

[0067] Based on the above embodiment, the information acquisition circuit module collects the load voltage of each output separately, and the main control board compares each load voltage with the corresponding target output voltage to close-loop adjust the sine wave pulse width ratio of the inverter module and the square wave pulse width ratio of the BUCK DC module to adjust their output voltage, so that they automatically track the power consumption of the equipment connected to the inverter module to quickly match the voltage level required by the user.

[0068] Based on the contents of the above embodiment, as an optional embodiment, the main control board is used to run a trained least squares support vector machine (LS-SVM) model; the LS-SVM model is used to perform a linear regression operation on the load voltage and the target output voltage to obtain fitting parameters; the main control board re-determines the square wave pulse width ratio of the BUCK DC module and the sine wave pulse width ratio of the inverter module based on the fitting parameters.

[0069] The main research focus for traditional photovoltaic systems is on improving the accuracy of maximum power control. A maximum operating point voltage prediction model based on the least squares method has been developed. This model predicts the maximum operating point voltage of the photovoltaic power generation system and uses the predicted voltage to correct the reference voltage of the constant voltage control method, thereby achieving maximum power tracking control of the photovoltaic power generation system with a relative error of approximately 0.1V. However, the least squares method is difficult to quickly and stably track the system's maximum power, and the control accuracy of the system's output voltage is low, especially given that the power supply has a constant voltage output.

[0070] However, the linear regression adopted by the least squares method is a regression analysis that models the relationship between one or more independent variables and dependent variables. Its characteristic is a linear combination of one or more model parameters called regression coefficients, as shown in expression (1).

[0071] f (x) =w1x1+w2x2+...+w n x n +b (1)

[0072] In general, there are the following problems:

[0073] (1) When the sample size is small, underfitting occurs, resulting in inaccurate functions and different errors in different level ranges.

[0074] (2) If the number of samples is increased, there are too many original features and some noisy features. The model is too complex because the model tries to take into account all the test data points, resulting in overfitting and making the function unprepared.

[0075] It can be seen that the traditional least squares regression operation sacrifices accuracy to meet the requirements of fast matching and complete system voltage tracking.

[0076] In view of this, the present invention separately collects the load voltages output by the BUCK DC module and the inverter module, and controls the square wave pulse width ratio of the BUCK DC module and the sine wave pulse width ratio of the inverter module through registers in the main control board, so as to achieve a regression logic operation between the actual output load voltage and the target output voltage set for the control output.

[0077] Figure 3 FIG. 1 is a flow chart of the voltage regression control provided by the present invention, as shown in FIG. Figure 3 As described above, after the gate dog is turned off, the crystal oscillator is configured, and the entire photovoltaic power supply device is system-formatted, the initial square wave pulse width ratio and initial sine wave pulse width ratio corresponding to the buck DC module and the inverter module are respectively set by a table lookup method, and the battery charging management module controls the buck DC module to output the corresponding initial direct current and initial alternating current according to this setting.

[0078] Furthermore, an information acquisition circuit module, such as an ADC circuit, collects and feeds back the load voltage to the main control board.

[0079] A pre-trained LS-SVM model is used to perform linear regression operations based on the load voltage and the corresponding target output voltage to obtain the corresponding fitting parameters.

[0080] The fitting parameters can be displayed by the host computer and fed back to the main control board, so that the main control board changes the counting memory, adjusts the corresponding pulse width ratio, and stores it as the initial pulse width ratio of the next output cycle.

[0081] In the next output cycle, the output voltages of the BUCK DC module and the inverter module are controlled again according to the new initial pulse width ratio, and the above steps are repeated.

[0082] The photovoltaic power supply device, DC-DC, and DC-AC provided by the present invention perform linear regression calculations on the load voltage and target output voltage of the buck DC module and the inverter module using the LS-SVM model. The fitting parameters are adjusted offline to achieve regulation of the square wave and sine wave of the buck DC module and the inverter module, making the system constant current output voltage range adjustable to meet the needs of low-power AC and DC devices.

[0083] Based on the content of the above embodiment, as an optional embodiment, the inverter module includes an IR2104 drive circuit, an LC low-pass filter circuit and a switch circuit;

[0084] The main control board outputs a bipolar sine wave to the control end of the IR2104 drive circuit through a table lookup method, and controls the cross conduction of the bipolar sine wave to control the switching frequency of the switching circuit;

[0085] The LC low-pass filter circuit is used to receive direct current input from the BUCK DC module and output alternating current;

[0086] The switching frequency of the switching circuit is used to adjust the output amplitude of the alternating current.

[0087] It should be noted that the present invention does not impose any specific limitation on the circuit structure of the inverter module, and the existing inverter voltage control circuit can be adopted.

[0088] Optionally, the main control board uses a table lookup method to output bipolar SPWM to reduce harmonic content, and the IR2104 drive circuit can be used to increase the SPWM drive power. A stable AC power is generated through LC low-pass filtering, and the final AC output amplitude is adjusted by the switching frequency of the switching circuit.

[0089] In addition, due to the characteristics of switching tubes (such as MOSFET, IGBT, etc.), there is a rise and fall time when they are turned on or off. In order to stagger the time of simultaneous opening, a dead zone is set to prevent the upper and lower bridge arms from being turned on at the same time (the upper and lower bridge arms being directly turned on is equivalent to the positive and negative poles being short-circuited). Therefore, the present invention controls the dead zone delay effect by cross-conducting SPWM in the inverter voltage control circuit.

[0090] The photovoltaic power supply device provided by the present invention has both AC and DC voltage outputs. While meeting the requirement of low power consumption, the power supply device outputs the maximum power of solar energy through a battery management system. At the same time, the AC voltage is protected by a dead-zone circuit to ensure that the photovoltaic power supply device performs voltage modulation under the drive of a bootstrap circuit, and short-circuiting of the modulated voltage will not occur, so that the system can output stable AC and DC voltages.

[0091] Based on the content of the above embodiment, as an optional embodiment, the control chip of the BOOST circuit module is a TPS61175 chip.

[0092] Figure 4 This is a circuit diagram of a battery charging management module provided by the present invention, such as Figure 4 As described above, the CN3722 module adopts a switching buck mode, realizes the maximum power point tracking function of the photovoltaic cell by means of DC-DC conversion and using a constant voltage method.

[0093] Figure 5 This is a circuit diagram of a BOOST circuit module provided by the present invention, such as Figure 5 As shown in the figure, the BOOST circuit module uses the TPS61175 as the BOOST boost circuit control chip. With an input of 5-20V, it can output up to 24V and a maximum drive current of 3A. The BOOST boost circuit composed of the TPS61175 can be used in a variety of standard switching regulator topologies, including boost, SEPIC, and flyback configurations.

[0094] Figure 6 This is a circuit diagram of a BUCK DC module provided by the present invention, such as Figure 6As shown in the figure, the BUCK DC module uses an H-bridge drive circuit. An 80kHz square wave signal is input to the PWM port, and the supply voltage is 24V, achieving an adjustable output between 0 and 24V. The output port is connected to an information acquisition circuit module. When the output voltage is too high or too low, the H-bridge PWM duty cycle is adjusted by controlling the registers on the main control board to achieve dynamic output voltage regulation.

[0095] Based on the content of the above embodiment, as an optional embodiment, it further includes a host computer, which is communicatively connected to the main control board;

[0096] The host computer is used to receive and store the historical load voltage of each of the BUCK DC modules and each of the inverter modules collected by the information collection circuit module, and to construct a training set and a test set;

[0097] The least squares support vector machine model is pre-trained using the training set and the test set.

[0098] Figure 7 This is a flowchart of a software and hardware platform for loading the LV-SVM model provided by the present invention. Figure 7 As shown, the photovoltaic power supply device provided by the present invention builds a software and hardware platform system. The host computer monitoring system monitors the AC and DC output voltages of the photovoltaic power supply device to meet the low-power power supply requirements of the equipment. The LV-SVM model is trained by collecting the pre-set target output voltages of each output interface and the collected load voltage until the trained model can quickly and accurately match the required voltage of the equipment.

[0099] Therefore, the photovoltaic power supply device provided by the present invention can adapt to multiple voltage levels in a step-by-step manner. In addition to having the characteristics of existing constant voltage output, it can meet the adaptation requirements of park equipment for multiple voltage levels. At the same time, under the high-precision and high-stability calculation of the regression method, this device can automatically track the 0-24V voltage of the equipment, and its response accuracy is 0.01V.

[0100] Figure 8 This is a schematic diagram of the photovoltaic power supply method provided by the present invention, such as Figure 8 As shown, it mainly includes the following steps:

[0101] Step S1, determining an initial square wave pulse width ratio of the BUCK DC module and an initial sine wave pulse width ratio of the inverter module, so that the battery charging management module controls the BUCK DC module to output a corresponding initial DC power, and controls the inverter module to output a corresponding initial AC power;

[0102] Step S2, receiving the load voltages output by the BUCK DC module and the inverter module collected by the information collection circuit module; the load voltages include the DC load voltage output by the BUCK DC module and the AC load voltage output by the inverter module;

[0103] Step S3: Sending the load voltage to the main control board, so that the main control board can redetermine a new square wave pulse width ratio of the buck DC module based on the DC load voltage and the target output voltage of the buck DC module; and redetermine a new sinusoidal wave pulse width ratio of the inverter module based on the AC load voltage and the target output voltage of the inverter module;

[0104] Step S4, feeding back the new square wave pulse width ratio and the new sine wave pulse width ratio to the battery charging management module, so that the battery charging management module adjusts the actual output voltages of the BUCK DC module and the inverter module respectively;

[0105] Step S5, iteratively execute steps S1 to S4.

[0106] As an optional embodiment, the main control board re-determines a new square wave pulse width ratio of the buck DC module based on the DC load voltage and the target output voltage of the buck DC module; and re-determines a new sinusoidal wave pulse width ratio of the inverter module based on the AC load voltage and the target output voltage of the inverter module, including:

[0107] Based on a pre-trained least squares support vector machine model, with the goal of controlling the photovoltaic cell to charge the battery pack in a maximum power point tracking manner, linear regression operations are performed on the DC load voltage and the AC load voltage with the corresponding target output voltage to obtain corresponding fitting parameters;

[0108] The square wave pulse width ratio of the BUCK DC module and the sine wave pulse width ratio of the inverter module are re-determined according to the fitting parameters.

[0109] It should be emphasized that the present invention also constructs training and test sets by increasing the number of data features of voltage levels;

[0110] Based on the k-fold cross validation method and while reducing the high-order effects, the least squares support vector machine model is pre-trained using the training set and the test set.

[0111] Unlike the linear regression used in the least squares method, machine learning approaches also face overfitting and underfitting in LS-SVM regression analysis. Overfitting (i.e., the model is too complex) occurs when a hypothesis fits the training data better than other hypotheses, but fails to fit the data well on datasets outside of the training data. Underfitting (i.e., the model is too simple) occurs when a hypothesis fails to fit the training data better, but fails to fit the data well on datasets outside of the training data.

[0112] Unlike existing photovoltaic power supply methods, which track output voltage using a traditional multi-order least squares mathematical model to achieve voltage regression calculations with an accuracy of approximately 0.1V, which sacrifices voltage output accuracy for fast matching, this solution is based on sufficient training of all voltage level data to obtain a stable voltage regression model with tracking accuracy within 0.01V.

[0113] Specifically, the present invention is based on a pre-trained LV-SVM model, and performs linear regression operations on the DC load voltage and the AC load voltage with the corresponding target output voltage to obtain corresponding fitting parameters.

[0114] Before using the LV-SVM model for actual calculation, the process of pre-collecting historical information and pre-training it is also included. Considering the occurrence of overfitting and underfitting that may occur during the pre-training process, the present invention adopts the following solution:

[0115] Solution (1): Increase the number of voltage level data features as the basis for feature selection and eliminate features with high correlation, for example:

[0116] All historical data (hereinafter referred to as samples) was divided into different voltage levels from 0 to 24V, with 3.3V, 5V, 12V, and 24V being primarily selected as data features. For the 3.3V voltage level, in voltage regression, new features were added to the data boundary at ±0.3, such as data greater than or equal to 3.27V and less than or equal to 3.33V. This increased the number of data features for the voltage level and eliminated features with high correlation. Furthermore, after data cleaning, the samples were further divided into training and test sets for model training.

[0117] Solution (2) Use k-fold cross-validation to tune the model. The trained model is then iterated again. During each iteration, each sample point has only one chance to be included in the training set or test set, rather than fixing the training set and test set.

[0118] Figure 9 This is a diagram showing the relative usage of the test set and training set in traditional model training, such as Figure 9 As shown in the figure, in traditional model training, the test set is a fixed set, which makes the training model prone to underfitting and overfitting under the influence of the number of samples.

[0119] Figure 10 Schematic diagram of the relative usage of the test set and the training set in the model training using k-fold cross validation provided by the present invention, such as Figure 10 As shown, for example, for a voltage level of 3.3V, all samples are divided into K sub-samples, a single sub-sample of which is retained as data for the validation model, and the other K-1 samples are used for training.

[0120] Cross-validation is repeated K times, with each subsample validated once, and the K results are averaged or combined to produce a single estimate. The advantage of this training method is that it uses randomly generated subsamples for training and validation, with each result validated once. Figure 11 This is a diagram of the training effect obtained by using different K values, such as Figure 11 As shown in the figure, by increasing the K value, the tuning effect of the LV-SVM model can be further optimized, and a smaller generalization error can be obtained.

[0121] Solution (3): Fast matching, reducing high-order effects, and making θ3 and θ4 infinitely close to 0;

[0122] θ0+θ1x+θ2x 2 +θ3x 3 +θ4x 4 (2)

[0123] Ignoring the influence of θ3 and θ4, the following is shown:

[0124] θ0+θ1x+θ2x 2 (3)

[0125] Increase the number of samples in the training set and test set so that samples of each voltage level are trained. The voltage levels cover 0 to 24V. Set each specific sample as a feature value, and select 3.3v, 5v, 12v, and 24v as specific samples to reduce high-order effects and ensure rapid system response.

[0126] The goal of the present invention is to find a linear function, such as equation (4):

[0127] f (x) =w T x+b (4)

[0128] The data collected by the host computer include the set voltage value and the output voltage value. If the data has a deviation of ±ε from the regression function, such as equations (5) and (6), it is constrained within the accuracy range.

[0129]

[0130]

[0131] However, during actual collection, there will still be some data scattered within ±ε, and it is necessary to introduce a relaxation factor and adopt a soft boundary method, such as constraint equation (7);

[0132] ζ i ,ζ i * ≥0 (7)

[0133] So there will be constraints as shown in equation (8):

[0134]

[0135] Figure 12 This is a diagram of the training and prediction results. Figure 13 is a schematic diagram of the test set prediction results, such as Figure 12 as well as Figure 13 As shown, the test set and training set of the sample are used with the set value of 0~24V (i.e. the target output voltage, also known as the label value), where mes and R 2 The test and training sets used to describe fitting accuracy and regression training are optimized through cross-validation while mitigating the influence of higher-order coefficients to obtain parameters with the smallest error from the optimal classification surface. The mathematical model is applied to the device to achieve both output voltage accuracy and optimal matching speed.

[0136] The photovoltaic power supply method provided by the present invention is different from traditional tracking algorithms. It focuses on the research of system stability output and adopts the LS-SVM method to reduce the influence of high-order terms to meet the demand for fast voltage matching. It also performs cross-validation of samples to ensure that all voltage level data are trained to meet the demand for high-precision voltage matching.

[0137] It should be noted that the photovoltaic power supply method provided by the present invention is implemented based on the photovoltaic power supply device described in any of the above embodiments during specific execution, which will not be described in detail in this embodiment.

[0138] Figure 14 Schematic diagram of the structure of the electronic device provided by the present invention, such as Figure 14As shown, the electronic device may include: a processor 410, a communication interface 420, a memory 430, and a communication bus 440, wherein the processor 410, the communication interface 420, and the memory 430 communicate with each other via the communication bus 440. The processor 410 may call the logic instructions in the memory 430 to execute the photovoltaic power supply method, which includes:

[0139] Step S1, determining an initial square wave pulse width ratio of the BUCK DC module and an initial sine wave pulse width ratio of the inverter module, so that the battery charging management module controls the BUCK DC module to output a corresponding initial DC power, and controls the inverter module to output a corresponding initial AC power;

[0140] Step S2, receiving the load voltages output by the BUCK DC module and the inverter module collected by the information collection circuit module; the load voltages include the DC load voltage output by the BUCK DC module and the AC load voltage output by the inverter module;

[0141] Step S3: Sending the load voltage to the main control board, so that the main control board can redetermine a new square wave pulse width ratio of the buck DC module based on the DC load voltage and the target output voltage of the buck DC module; and redetermine a new sinusoidal wave pulse width ratio of the inverter module based on the AC load voltage and the target output voltage of the inverter module;

[0142] Step S4, feeding back the new square wave pulse width ratio and the new sine wave pulse width ratio to the battery charging management module, so that the battery charging management module adjusts the actual output voltages of the BUCK DC module and the inverter module respectively;

[0143] Step S5, iteratively execute steps S1 to S4.

[0144] In addition, the logic instructions in the above-mentioned memory 430 can be implemented in the form of a software functional unit and can be stored in a computer-readable storage medium when sold or used as an independent product. Based on this understanding, the technical solution of the present invention, or the part that contributes to the prior art, or the part of the technical solution, can be embodied in the form of a software product. The computer software product is stored in a storage medium and includes several instructions for enabling a computer device (which can be a personal computer, server, or network device, etc.) to execute all or part of the steps of the device described in each embodiment of the present invention. The aforementioned storage medium includes: various media that can store program codes, such as a USB flash drive, a mobile hard disk, a read-only memory (ROM), a random access memory (RAM), a magnetic disk or an optical disk.

[0145] On the other hand, the present invention further provides a computer program product, comprising a computer program stored on a non-transitory computer-readable storage medium, wherein the computer program comprises program instructions. When the program instructions are executed by a computer, the computer is capable of performing the photovoltaic power supply method provided by each of the above-mentioned devices, the method comprising:

[0146] Step S1, determining an initial square wave pulse width ratio of the BUCK DC module and an initial sine wave pulse width ratio of the inverter module, so that the battery charging management module controls the BUCK DC module to output a corresponding initial DC power, and controls the inverter module to output a corresponding initial AC power;

[0147] Step S2, receiving the load voltages output by the BUCK DC module and the inverter module collected by the information collection circuit module; the load voltages include the DC load voltage output by the BUCK DC module and the AC load voltage output by the inverter module;

[0148] Step S3: Sending the load voltage to the main control board, so that the main control board can redetermine a new square wave pulse width ratio of the buck DC module based on the DC load voltage and the target output voltage of the buck DC module; and redetermine a new sinusoidal wave pulse width ratio of the inverter module based on the AC load voltage and the target output voltage of the inverter module;

[0149] Step S4, feeding back the new square wave pulse width ratio and the new sine wave pulse width ratio to the battery charging management module, so that the battery charging management module adjusts the actual output voltages of the BUCK DC module and the inverter module respectively;

[0150] Step S5, iteratively execute steps S1 to S4.

[0151] In another aspect, the present invention further provides a non-transitory computer-readable storage medium having a computer program stored thereon. When the computer program is executed by a processor, the photovoltaic power supply method provided in each of the above embodiments is implemented. The method includes:

[0152] Step S1, determining an initial square wave pulse width ratio of the BUCK DC module and an initial sine wave pulse width ratio of the inverter module, so that the battery charging management module controls the BUCK DC module to output a corresponding initial DC power, and controls the inverter module to output a corresponding initial AC power;

[0153] Step S2, receiving the load voltages output by the BUCK DC module and the inverter module collected by the information collection circuit module; the load voltages include the DC load voltage output by the BUCK DC module and the AC load voltage output by the inverter module;

[0154] Step S3: Sending the load voltage to the main control board, so that the main control board can redetermine a new square wave pulse width ratio of the buck DC module based on the DC load voltage and the target output voltage of the buck DC module; and redetermine a new sinusoidal wave pulse width ratio of the inverter module based on the AC load voltage and the target output voltage of the inverter module;

[0155] Step S4, feeding back the new square wave pulse width ratio and the new sine wave pulse width ratio to the battery charging management module, so that the battery charging management module adjusts the actual output voltages of the BUCK DC module and the inverter module respectively;

[0156] Step S5, iteratively execute steps S1 to S4.

[0157] The method embodiments described above are merely illustrative. The units described as separate components may or may not be physically separate, and the components shown as units may or may not be physical units, i.e., they may be located in one location or distributed across multiple network elements. Some or all of the modules may be selected based on actual needs to achieve the objectives of the present embodiment. Persons of ordinary skill in the art will be able to understand and implement the present invention without inventive effort.

[0158] Through the description of the above embodiments, those skilled in the art can clearly understand that each embodiment can be implemented by means of software plus the necessary general hardware platform, or of course, by hardware. Based on this understanding, the essence of the above technical solution or the part that contributes to the existing technology can be embodied in the form of a software product. The computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, a magnetic disk, an optical disk, etc., and includes a number of instructions for enabling a computer device (which can be a personal computer, a server, or a network device, etc.) to execute the apparatus described in each embodiment or certain parts of the embodiment.

[0159] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the various embodiments of the present invention.

Claims

1. A photovoltaic power supply device, characterized in that: include: Photovoltaic cells, battery charging management module, battery pack, BOOST circuit module, main control board, information acquisition circuit module, BUCK DC module and inverter module; The battery charging management module is used to control the photovoltaic cell to charge the battery pack in a maximum power point tracking manner; The input end of the BOOST circuit module is connected to the output end of the battery pack; The BUCK DC module and the inverter module are respectively connected to the output end of the BOOST circuit module; The information acquisition circuit module is used to collect the load voltage of the BUCK DC module and the inverter module, and feed the load voltage back to the main control board; The main control board re-determines the square wave pulse width ratio of the BUCK DC module and the sine wave pulse width ratio of the inverter module according to the load voltage and the target output voltages of the BUCK DC module and the inverter module, and feeds back the information to the battery charging management module; The battery charging management module adjusts the actual output voltages of the BUCK DC module and the inverter module respectively according to the square wave pulse width ratio and the sine wave pulse width ratio; The main control board is used to run the trained least squares support vector machine model; The least squares support vector machine model is used to perform a linear regression operation on the load voltage and the target output voltage to obtain fitting parameters; The main control board re-determines the square wave pulse width ratio of the BUCK DC module and the sine wave pulse width ratio of the inverter module according to the fitting parameters.

2. The photovoltaic power supply device according to claim 1, characterized in that: The inverter module includes an IR2104 drive circuit, an LC low-pass filter circuit and a switch circuit; The main control board outputs a bipolar sine wave to the control end of the IR2104 drive circuit through a table lookup method, and controls the cross conduction of the bipolar sine wave to control the switching frequency of the switching circuit; The LC low-pass filter circuit is used to receive direct current input from the BUCK DC module and output alternating current; The switching frequency of the switching circuit is used to adjust the output amplitude of the alternating current.

3. The photovoltaic power supply device according to claim 1, characterized in that: The control chip of the BOOST circuit module is the TPS61175 chip.

4. The photovoltaic power supply device according to claim 1, characterized in that: The BUCK DC module includes an H-bridge drive circuit.

5. The photovoltaic power supply device according to claim 1, characterized in that: There are multiple BUCK DC modules and / or multiple inverter modules.

6. The photovoltaic power supply device according to claim 1, characterized in that: It also includes a host computer, which is communicatively connected to the main control board; The host computer is used to receive and store the historical load voltage of each of the BUCK DC modules and each of the inverter modules collected by the information collection circuit module, and to construct a training set and a test set; The least squares support vector machine model is pre-trained using the training set and the test set.

7. A photovoltaic power supply method implemented based on the photovoltaic power supply device according to any one of claims 1 to 6, characterized in that: include: Step S1, determining an initial square wave pulse width ratio of the BUCK DC module and an initial sine wave pulse width ratio of the inverter module, so that the battery charging management module controls the BUCK DC module to output a corresponding initial DC power, and controls the inverter module to output a corresponding initial AC power; Step S2, receiving the load voltages output by the BUCK DC module and the inverter module collected by the information collection circuit module; the load voltages include the DC load voltage output by the BUCK DC module and the AC load voltage output by the inverter module; Step S3: Sending the load voltage to the main control board, so that the main control board can redetermine a new square wave pulse width ratio of the buck DC module based on the DC load voltage and the target output voltage of the buck DC module; and redetermine a new sinusoidal wave pulse width ratio of the inverter module based on the AC load voltage and the target output voltage of the inverter module; Step S4, feeding back the new square wave pulse width ratio and the new sine wave pulse width ratio to the battery charging management module, so that the battery charging management module adjusts the actual output voltages of the BUCK DC module and the inverter module respectively; Step S5, iteratively execute steps S1 to S4.

8. The photovoltaic power supply method according to claim 7, characterized in that: The main control board re-determines a new square wave pulse width ratio of the BUCK DC module according to the DC load voltage and the target output voltage of the BUCK DC module; and re-determining a new sinusoidal wave pulse width ratio of the inverter module according to the AC load voltage and the target output voltage of the inverter module, comprising: Based on a pre-trained least squares support vector machine model, with the goal of controlling the photovoltaic cell to charge the battery pack in a maximum power point tracking manner, linear regression operations are performed on the DC load voltage and the AC load voltage with the corresponding target output voltage to obtain corresponding fitting parameters; The square wave pulse width ratio of the BUCK DC module and the sine wave pulse width ratio of the inverter module are re-determined according to the fitting parameters.

9. The photovoltaic power supply method according to claim 8, characterized in that: Also includes: Increase the number of voltage level data features to construct training and test sets; Based on the k-fold cross validation method and while reducing the high-order effects, the least squares support vector machine model is pre-trained using the training set and the test set.

Images