Photovoltaic mppt dual mode control system and method

Abstract

The application discloses a photovoltaic MPPT dual-mode control system and method, which comprises a photovoltaic component unit, a Boost circuit, a voltage and current sampling circuit, an MPPT controller, a PWM modulation module and a direct-current electronic load. The MPPT controller adopts a dual-mode power prediction control strategy combining an exponential adjustable step size conductance increment method and a constant voltage tracking method, a disturbance trend prediction quantity fusing first-order and second-order power changes is constructed, and forward-looking identification of environmental changes is realized; the width of a hysteresis interval of mode switching is dynamically adjusted based on output power fluctuation variance, and noise resistance is improved; in the conductance increment mode, an exponential adaptive step size is generated by using a nonlinear potential function containing a first-order slope and a second-order curvature, and high-precision control of 'fast tracking with a large step size far from the MPP and small step size to suppress oscillation near the MPP' is realized. The tracking speed and the steady-state precision of the system under dynamic and steady-state conditions are effectively improved, and the overall energy capture efficiency is enhanced.

Description

TECHNICAL FIELD

[0001] The application belongs to the technical field of photovoltaic power generation control, and particularly relates to a dual-mode maximum power point tracking (MPPT) control system fusing trend prediction, adaptive hysteresis judgment and nonlinear potential function exponential step length adaptive mechanism, and suitable for a photovoltaic power generation system containing a Boost converter. BACKGROUND

[0002] With large-scale application of photovoltaic power generation in the field of distributed power supply and new energy, power output of the photovoltaic power generation system is significantly random and time-varying due to influences of environmental irradiance, temperature and partial shading, so that the maximum power point (MPP) of the photovoltaic system dynamically changes. Performance of the MPPT technology directly determines energy utilization efficiency of the photovoltaic system.

[0003] The traditional incremental conductance method (INC) is widely used due to simple implementation, but a fixed step length is generally used, and problems such as slow dynamic tracking speed caused by small step length, significant steady-state oscillation caused by large step length and obvious response delay when facing irradiance mutation exist. In order to improve the problems, methods such as variable step length INC, adaptive LMS step length and exponential step length adjustment are introduced. However, there are still problems such as single step length adjustment dimension, linear / nonlinear mapping of only first-order power change or voltage and current change at adjacent sampling times , difficulty in considering rapidity and steady-state accuracy, and single structure of the known exponential step length function, which is easy to cause step length overshoot in a noise or large curvature area. The current mode switching strategy generally uses a fixed threshold, cannot process jitter under noise conditions, and cannot perceive disturbance trend in advance, and can only respond in a lagging manner.

[0004] Therefore, a new type of MPPT technology is needed, which can realize fast tracking in dynamic changes, inhibit oscillation under steady-state conditions, predict disturbance trend in advance to improve accuracy of mode switching, and has a high-performance intelligent MPPT control method with multi-dimensional distance measurement and stable step length adjustment mechanism. SUMMARY

[0005] The application proposes a dual-mode MPPT control system based on trend prediction, adaptive hysteresis judgment mechanism and nonlinear potential function exponential step length adjustment, aiming at problems such as slow maximum power point tracking (MPPT) response speed, large steady-state oscillation, untimely mode switching and noise interference sensitivity of the photovoltaic power generation system under complex working conditions such as dynamic irradiance, temperature fluctuation and load disturbance. The application realizes comprehensive performance improvement of the photovoltaic array under dynamic conditions in terms of forward-looking prediction, robust mode switching and high-precision steady-state tracking by constructing a three-layer intelligent control framework of “trend prediction—mode determination—adaptive step length”.

[0006] To achieve the above-mentioned purpose of the application, the technical scheme provided by the application is as follows:

[0007] A photovoltaic MPPT dual-mode control system, comprising a photovoltaic component unit, a Boost circuit, a voltage sampling circuit, a current sampling circuit, an MPPT controller, a PWM pulse width modulation module and a direct current electronic load; the photovoltaic component unit is connected with the Boost circuit, the voltage sampling circuit and the current sampling circuit sample the voltage signal and the current signal generated by the photovoltaic component unit respectively and then send them to the MPPT controller, the MPPT controller uses the dual-mode power prediction control method combining the conductance increment method with the constant voltage tracking method to automatically adjust the duty cycle, then sends the optimal duty cycle signal to the PWM pulse width modulation module, and finally outputs the PWM wave to the power switch tube IGBT in the Boost circuit through the PWM pulse width modulation module, so as to control the on-off of the power switch tube IGBT and stabilize the output power of the photovoltaic component unit at the maximum value.

[0008] The photovoltaic component unit comprises a plurality of photovoltaic cells, the Boost circuit comprises a capacitor C1, a capacitor C2, an inductor L, a diode D and a power switch tube IGBT, the photovoltaic component unit generates positive and negative electrode voltage at the two ends of the energy storage element capacitor C1 after being connected into the Boost circuit, the positive electrode of the photovoltaic component unit is connected with the inductor L, the other end of the inductor L is connected with the collector of the power switch tube IGBT and the forward conduction end of the diode D, the cathode of the diode D is connected with one end of the capacitor C2 and the direct current electronic load, the gate of the power switch tube IGBT is connected with the output end of the PWM pulse width modulation module, and the emitter of the power switch tube IGBT is connected with the negative electrode of the photovoltaic component unit.

[0009] The sampled voltage signal is derived from the signal processed by the voltage dividing resistor at the input end of the Boost circuit and is connected with the input end of the MPPT controller after proportional linear correction by the voltage sensor.

[0010] The sampling position of the current signal is at the connection between the output end of the photovoltaic component unit and the capacitor C1, and the sampled current signal is connected with the input end of the MPPT controller after proportional linear correction by the current sensor.

[0011] The application further discloses a control method of the photovoltaic MPPT dual-mode control system, comprising the following steps:

[0012] Step 1: periodically collecting the output voltage and the output current of the photovoltaic component unit and calculating the output power.

[0013] Step 2: using the first-order change of power and the second-order change of power to construct the disturbance trend prediction quantity , for early identification of environmental change trends. Disturbance trend prediction Not only reflects the speed of illumination change, but also reflects its acceleration, so that the system can identify the upcoming light sudden change in advance and make forward-looking control.

[0014] Step 3: Dynamically adjust the hysteresis interval width of mode switching according to the fluctuation variance of output power, realize noise adaptive mode judgment, and based on disturbance trend prediction With noise adaptive hysteresis judgment result, select conductance increment mode or constant voltage tracking mode. The present application introduces a noise perception mechanism based on the fluctuation variance of output power Real-time adjustment of hysteresis interval width , solves the problem of fixed threshold switching easily disturbed by noise and frequent jumps, and realizes stable and reliable mode determination.

[0015] Step 4: In constant voltage tracking mode, set reference voltage based on open circuit voltage ratio, and adjust duty ratio through PI controller.

[0016] Step 5: In conductance increment mode, generate adaptive step based on nonlinear potential function, and update duty ratio.

[0017] Firstly, based on the geometric characteristics of photovoltaic power-voltage curve, a nonlinear potential function containing first-order slope and second-order curvature is proposed To more accurately reflect the distance between the current working point and the maximum power point MPP, significantly improve the defects of traditional Measurements such as instability in large curvature area and insufficient resolution in flat area. Then, based on the nonlinear potential function Design of exponential decay step Make the step increase when far away from the maximum power point MPP to quickly approach, and automatically shrink when close to the maximum power point MPP to realize low-oscillation accurate tracking, with natural nonlinear adaptive characteristics.

[0018] Step 6: Dual-mode cooperative mechanism: combine the improved conductance increment mode of step 5 with the constant voltage tracking mode of step 4, and realize intelligent selection of mode through disturbance trend prediction And noise adaptive hysteresis judgment result, realize intelligent selection of mode, so that the system can automatically maintain the optimal control mode under different working conditions.

[0019] Step 7: Generate PWM wave according to the final duty ratio to control the on-off of power switch tube in Boost circuit, realize maximum power point tracking.

[0020] Further, the step 1 includes the following steps:

[0021] The controller periodically collects the output voltage And output current of photovoltaic module Calculate the output power The changes in voltage and current at adjacent sampling times are defined as follows:

[0022]

[0023] in, The sampling period is for The output voltage is collected at all times. for The output voltage is collected at all times. This refers to the change in voltage. for The output current is collected at all times. for The output current is collected at all times. This represents the change in current.

[0024] The power change rate and derivative ratio are calculated based on the changes in voltage and current to determine the position of the operating point relative to the maximum power point (MPP).

[0025] Furthermore, step 2 includes the following steps:

[0026] After collecting the output voltage and current of the photovoltaic module, the data enters the mode determination module, which introduces short-term disturbance trend prediction. It can identify disturbance trends in advance and switch to a suitable mode before environmental changes occur, greatly improving the dynamic response capability of the MPPT algorithm.

[0027]

[0028] in, This represents the power change during the current sampling period. The second-order time difference of the output power is used to reflect the acceleration of the power change. and The weighting coefficients were determined experimentally and were set to 1 and 0.4, respectively.

[0029] This perturbation trend prediction is mainly used to switch modes in advance and improve the dynamic response capability of the MPPT algorithm.

[0030] when When the system is about to experience a sudden change in light or temperature, it will enter the incremental conductivity mode (INC mode) in advance.

[0031] when When the system reaches a certain point, it is determined that it will stabilize and enter constant voltage tracking mode (CV mode).

[0032] in is a prediction threshold, determined by an adaptive hysteresis interval, which solves the problem of frequent switching in current technology.

[0033] Further, the step 3 comprises the following steps:

[0034] Calculate the noise estimate of the output signal (fluctuation variance) :

[0035]

[0036] wherein, is a function of the calculated variance, is the output power at time k, is the short-term power average.

[0037] Based on Adjust the hysteresis interval width :

[0038]

[0039] wherein, is the dynamic hysteresis width, is the base width, is the weight factor.

[0040] At this time, the entry condition of the INC mode is:

[0041]

[0042] The entry condition of the CV mode is:

[0043]

[0044] wherein, is a reference threshold, usually 0.

[0045] The mode determination module adopts an adaptive hysteresis interval to avoid frequent switching. The greater the noise, the wider the hysteresis interval. Conversely, the more stable the environment, the narrower the hysteresis interval, thereby improving the adaptability of control.

[0046] Further, the step 4 comprises the following steps:

[0047] When the environment is stable and the MPPT controller switches to the constant voltage tracking mode, the reference voltage is taken as a certain proportion of the open-circuit voltage of the photovoltaic module:

[0048]

[0049] The value of is selected according to the actual parameters of the photovoltaic module.​

[0050] The update rule of constant voltage tracking method duty cycle is:

[0051]

[0052] The voltage deviation is ; the proportional gain reflects the "strength" of the response to the current voltage error, the larger the system response is faster, the stronger the ability to correct errors; integral gain is used to eliminate steady-state error. It will accumulate the historical error, as long as there is a small error, the integral term will continue to act until the error is zero.

[0053] In order to design the PI controller, its working point near the maximum power point is linearized and simplified as a first-order inertia plus pure delay system:

[0054]

[0055] Among them, is the system gain, is the time constant of the system, is the delay of the system, mainly from the sampling, calculation and PWM update of digital control, is the delay transfer function.

[0056] The gain can be estimated as the order of magnitude, according to the calculation ; the time constant is mainly determined by the input capacitance of the Boost circuit and the system impedance. Usually small, according to the circuit parameters to take 5 ms; delay For digital control system (DSP), the delay is usually 1~2 sampling periods. The sampling frequency is 10 kHz, so take .

[0057] Because is small, the simple Ziegler-Nichols first method (step response method) can be used to realize the fast and small overshoot or no overshoot of CV mode.

[0058]

[0059]

[0060] After calculation, , . Through the smooth adjustment of PI control loop, it is ensured that it converges quickly and stably when the light is stable.

[0061] Further, the step 5 comprises the following steps:

[0062] When the environmental disturbance is large, the MPPT controller switches to the improved conductance incremental method mode, and the dynamic tracking of the maximum output power of the photovoltaic module is realized through the improved INC algorithm with adaptive step size.

[0063] Step 5.1: In order to accurately and robustly measure the "distance" between the current operating point and the maximum power point (MPP), the present application proposes a nonlinear potential function based on the first-order slope and the second-order curvature of the power-voltage curve The potential function aims to reflect the direction and size of power change (information about approaching or moving away from the MPP) and the influence of curve shape (inflection point / curvature) at the same time, so as to provide a physically meaningful and noise-robust measurement input for step size adjustment. Unlike the method of using only, the potential function automatically gives greater inhibition to the area with significant curvature, so that the step size naturally decreases near the inflection point, avoiding overshooting and oscillation; in the flat area of the curve, it maintains higher sensitivity to speed up tracking.

[0064] In continuous time, the nonlinear potential function is defined as:

[0065]

[0066] Where is the first-order derivative of power with respect to voltage (slope); is the second-order derivative of power with respect to voltage (curvature); is the curvature weight coefficient, used to adjust the influence strength of curvature on the potential function, which is taken as 0.2 here, which can achieve the best balance between suppressing oscillation in the curvature area and maintaining sensitivity in the flat area.

[0067] In the controller, the first and second-order derivatives need to be estimated from discrete sampling data, and the following steps are adopted to balance response speed and noise resistance.

[0068] First, estimate the first-order derivative, and the discrete data ,

[0069]

[0070] To reduce noise, first do the exponential moving average (EMA) on P and V:

[0071]

[0072]

[0073] Where the smoothing coefficient 0.05 is determined by experiment.

[0074] The difference is calculated based on the smoothed data:

[0075]

[0076] Then the second derivative is estimated, which can be approximated by the second difference:

[0077]

[0078] wherein, is a small constant to prevent division by zero.

[0079] In order to avoid misleading second-order values calculated under the dominance of noise, the present application calculates the variance of the slope sequence and replaces the original second-order value with a confidence :

[0080]

[0081] wherein is a constant, which is 100 here. Then the is used instead of the original second-order value into the potential function. This strategy automatically reduces the second-order contribution under high noise and increases robustness.

[0082] Step 5.2: The present application maps the potential function Φ into an exponential adaptive step size μ to realize the adaptive law of "far big and near small". The basic mapping relationship is:

[0083]

[0084] wherein is the minimum step size, which ensures that the step size is not 0 and maintains the minimum disturbance ability; is the maximum step size, which ensures the rapid search ability, and is determined by the system dimension and sampling rate and are 0.002 and 0.12 respectively. The mapping relationship adopts an exponential relationship , when , the power output is closest to the MPP, the exponential term tends to 1, and the step size tends to the minimum step size; when , the power output is far from the MPP, the exponential term tends to 0, and the step size tends to the maximum step size, quickly adjusting the power output of the photovoltaic module. The square term enhances the suppression sensitivity to , and can respond more timely to the distance between the output power and the MPP.

[0085] In order to avoid step size mutation, the present application performs EMA smoothing on :

[0086]

[0087] in The step size smoothing coefficient was determined to be 0.15 through experiments.

[0088] At the same time Apply soft limiting:

[0089]

[0090] in, For the amplitude limiting function, This is for outputting the step size.

[0091] Furthermore, step 7 includes the following steps:

[0092] After receiving the step size update, the optimal duty cycle signal is sent to the PWM pulse width modulation module. Finally, the PWM pulse width modulation module outputs a PWM wave to the power switching transistor IGBT in the Boost circuit, so as to realize the photovoltaic module unit to continuously and stably output the maximum power.

[0093] Compared with the prior art, the beneficial effects of the present invention are as follows:

[0094] 1. Forward-looking trend prediction: By constructing the first-order change of fusion power With second-order changes Disturbance trend forecast The system can identify the "trend" and "acceleration" of sudden changes in light or temperature in advance, enabling a mode switch from "lagging response" to "advance prediction". This greatly shortens the response time under dynamic disturbances, allowing the system to lock onto the new maximum power point (MPP) more quickly.

[0095] 2. Precise distance measurement using nonlinear potential functions: The proposed potential function... It not only includes the traditional first-order slope It also innovatively introduced second-order curvature. As a compensation factor, this makes the system's measurement of the "distance" between the operating point and the MPP more accurate and robust, especially in areas with large curvature of the PV curve, effectively preventing misjudgments and providing a more reliable basis for step size adjustment.

[0096] 3. Exponential adaptive step size mechanism: based on potential function Designed exponential decay step size This achieves nonlinear adaptive adjustment of "large step size for rapid approximation at far MPP, and small step size for fine tracking at near MPP". When approaching MPP, the step size automatically shrinks to a minimum value, significantly reducing power oscillations in steady state, reducing power loss, and thus improving the average energy capture efficiency of the system.

[0097] 4. Noise-adaptive hysteresis switching mechanism: Introduces a mechanism based on output power fluctuation variance. dynamic hysteresis interval width When high noise is detected, the hysteresis interval for mode switching is automatically widened, effectively avoiding frequent mode jumps and malfunctions caused by noise; when the environment is stable, the hysteresis interval is narrowed to ensure switching sensitivity. This makes the control system highly robust to measurement noise and external interference. Attached Figure Description

[0098] Figure 1 This is a schematic diagram of the photovoltaic MPPT dual-mode control system provided by the present invention;

[0099] Figure 2 This is a simulation model diagram of the photovoltaic system of the present invention;

[0100] Figure 3 This is a simulation model diagram of the dual-mode power prediction algorithm of the present invention;

[0101] Figure 4 This is a schematic diagram of the MPPT controller of the present invention;

[0102] Figure 5 The output power diagrams of photovoltaic systems using the traditional incremental conductivity method and the method of this invention are shown.

[0103] Figure 6 This is a performance comparison chart between the traditional incremental conductivity method and the method of this invention. Detailed Implementation

[0104] The present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments. These embodiments are implemented based on the technical solution of the present invention, but the scope of protection of the present invention is not limited to the following embodiments.

[0105] This invention provides a high-performance dual-mode maximum power point tracking control system based on trend prediction and adaptive potential function. (See attached diagram) Figure 1 As shown, the system mainly includes: photovoltaic module unit 1, Boost circuit 2, voltage sampling circuit 3, current sampling circuit 4, MPPT controller 5, PWM pulse width modulation module 6, and DC electronic load 7.

[0106] The photovoltaic module unit 1 is composed of a plurality of photovoltaic cells connected in series and parallel, and the output positive and negative electrodes thereof are connected with the input end of the Boost circuit 2. The Boost circuit 2 is a DC-DC converter, and the core elements thereof include an input capacitor C1, an output capacitor C2, an energy storage inductor L, a power diode D, and a power switch tube IGBT as a switching device. The specific connection relationship is as follows: the positive electrode of the photovoltaic module unit 1 is connected with one end of the energy storage inductor L, the other end of the energy storage inductor L is connected with the collector of the power switch tube IGBT and the anode of the power diode D at the same time; the cathode of the power diode D is connected with one end of the output capacitor C2 and a direct current electronic load 7; the emitter of the IGBT, the other end of the input capacitor C1 and the output capacitor C2, and the other end of the direct current electronic load 7 are commonly connected to the negative electrode of the photovoltaic module unit 1, thereby forming a loop. The gate of the power switch tube IGBT receives a PWM signal generated from a PWM pulse width modulation module 6, and the duty cycle of the Boost circuit 2 is adjusted by controlling the on-off of the PWM signal, so as to change the operating point of the photovoltaic module.

[0107] The photovoltaic module unit 1 is configured as follows: the number of parallel strings: 40; the number of modules connected in series in each string: 10; the parameters of a single photovoltaic module: maximum power: 213.15 W, the number of cell pieces: 60, open circuit voltage : 36.3 V, short circuit current : 7.84 A, maximum power point voltage : 29 V, maximum power point current : 7.35 A, temperature coefficient : -0.36099% / ℃, and : 0.102% / ℃. It is calculated that the total open circuit voltage is 363 V, the total short circuit current is 313.6 A, the total maximum power point voltage is 290 V, the total maximum power point current is 294 A, and the maximum output power is 85.26 W.

[0108] The Boost circuit 2 is designed based on the photovoltaic array power of 85.26 kW. The input capacitor C1 is 1000 uF, the output capacitor C2 is 680 uF, the energy storage inductor L is 300 uH, the power switch tube is an ideal IGBT switch module, the power diode D is an ideal diode module, the switching frequency is 10 kHz, and the load resistance is valued at 700 V according to the typical bus voltage, and the resistance is 5.74 Ω.

[0109] The voltage sampling circuit 3 adopts a high-precision resistance voltage dividing network to sample the input end voltage (i.e. the terminal voltage of the photovoltaic module) of the Boost circuit 2 , and after proportional linear correction and isolation by a voltage sensor (such as an isolation operational amplifier), the voltage is sent to the ADC interface of the MPPT controller 5.

[0110] The current sampling circuit 4 employs a Hall current sensor, which is connected in series between the positive output of the photovoltaic module unit 1 and the input capacitor C1, to sample the photovoltaic output current . The sampling signal is sent to the ADC interface of the MPPT controller 5 after linear correction.

[0111] The MPPT controller 5 is the core of the system, preferably a digital signal processor (DSP) or a high-performance microcontroller (MCU). It periodically collects and , executes the dual-mode MPPT algorithm proposed in the present application, calculates the optimal duty ratio, and outputs the digital signal to the PWM pulse width modulation module 6.

[0112] As shown in Figure 4 , the core control algorithm of the present application is implemented in the MPPT controller 5 according to the following steps:

[0113] Step 1: Data acquisition and preprocessing:

[0114] The controller collects the photovoltaic terminal voltage and the output current in each sampling period , and calculates the instantaneous output power .

[0115] To suppress sampling noise, the voltage and power data are subjected to exponential moving average (EMA) filtering:

[0116]

[0117] where the smoothing coefficient is preferably 0.05. and are the smoothed data used for subsequent calculations.

[0118] Step 2: Trend prediction and mode determination

[0119] Calculate the trend prediction : To achieve early perception of environmental disturbances, the controller calculates a trend prediction that combines the power change rate and acceleration:

[0120]

[0121] where is the first-order difference of power, is the second-order difference of power (acceleration). The weight coefficient is optimized through experiments and takes , .

[0122] Adaptive hysteresis band calculation: To solve the problem of frequent switching under noise with fixed threshold, dynamic hysteresis band is introduced. First, estimate the noise level of the current system:

[0123]

[0124] where, is the function of variance calculation, is the short-time power average. Then, according to the noise level, the hysteresis band width is dynamically adjusted :

[0125]

[0126] where, the base width and the weight factor are pre-set according to the system tolerance.

[0127] Mode switching logic: The system sets a reference threshold (preferably 0) and generates a dynamic switching threshold:

[0128]

[0129] When , it is determined as a dynamic working condition, and the system switches to the improved incremental conductance mode (INC mode) in advance.

[0130] When , it is determined as a steady-state working condition, and the system switches to the constant voltage tracking mode (CV mode).

[0131] When , the current mode remains unchanged. This mechanism ensures the robustness of mode switching.

[0132] Step three: dual-mode control execution:

[0133] A. Constant voltage tracking mode (CV mode) implementation

[0134] When the system enters the CV mode, the reference voltage is set as:

[0135]

[0136] where, is the open-circuit voltage of the photovoltaic module, and the coefficient is selected between 0.75 and 0.82 according to the characteristics of the photovoltaic cell. A PI controller is used to adjust the duty cycle to make track . The discrete form of the PI controller is:

[0137]

[0138] where the voltage error is .

[0139] PI parameters are tuned based on the system model. The system in CV mode is linearized around MPP as a first order inertia plus pure time delay system . Where the system gain is estimated as O(33), and the time constant is determined by the circuit parameters, taking 5ms, and the delay is determined by the digital control period, taking 0.2ms when the sampling frequency is 10kHz. The Ziegler-Nichols first method is used to tune the PI parameters:

[0140]

[0141]

[0142] This parameter ensures fast and undamped response in CV mode.

[0143] B. Improved incremental conductance mode (INC mode) implementation

[0144] When the system enters INC mode, an adaptive step size mechanism based on a nonlinear potential function is adopted.

[0145] Compute the nonlinear potential function : To accurately measure the distance between the operating point and MPP, define the potential function:

[0146]

[0147] First derivative estimation: Compute using preprocessed data, where is a small constant to prevent division by zero.

[0148] Second derivative estimation and confidence weighting: To enhance robustness, first compute the variance of the first derivative sequence , then compute the confidence , with constant taking 100. The second derivative estimate is weighted by the confidence and used as: . The curvature weight coefficient takes 0.2.

[0149] Generate the exponential adaptive step size : Map the potential function to the step size:

[0150]

[0151] where, and are the maximum and minimum step size, respectively, which are taken as 0.12 and 0.002 in this embodiment. The exponential function ensures that the step size increases with increasing distance from the maximum power point, achieving the "far large and near small".

[0152] Step size smoothing and clipping: To prevent sudden changes in step size, the step size is smoothed by an EMA:

[0153]

[0154] The smoothing coefficient is taken as 0.15. Finally, the smoothed step size is soft-clipped:

[0155] .

[0156] Duty cycle update: In the INC mode, the duty cycle is updated according to:

[0157]

[0158] where, is the duty cycle at time k+1, is the duty cycle at time k. The direction of increase or decrease is determined by the sign of the basic criterion .

[0159] Step four: PWM generation and power device driving:

[0160] Regardless of the mode, the final duty cycle signal calculated by the MPPT controller 5 will be sent to the PWM pulse width modulation module 6. This module generates a high-frequency PWM wave with a corresponding duty cycle according to the duty cycle signal and applies it to the gate of the IGBT in the Boost circuit 2 through the driving circuit, controlling its on-off, thereby dynamically adjusting the operating point of the photovoltaic module to stabilize it near the maximum power point.

[0161] Test condition setting:

[0162] As shown in Figure 2 and Figure 3 , the simulation duration is set to 5s.

[0163] Irradiance variation: 0-1.0s is 600W / m 2 linearly rising to 800 W / m 2 , simulating the morning irradiance enhancement; 1.0-2.0s suddenly drops from 800W / m 2 to 400W / m 2 ​, the cloud cover, rapid decline; 2.0-3.0s from 400W / m 2 sudden increase to 900W / m 2 , the cloud over, the radiation recovery, the radiation intensity sudden increase; 3.0-4.0s linear decline from 900W / m 2 to 700W / m 2 ; 4.0-5.0s simulation thin cloud fluctuation, 700W / m 2 in W / m 2 radiation range fluctuation.

[0164] Temperature change: 0-1.0s temperature from 20℃ linearly rises to 25℃; 1.0-2.0s remains 25℃ unchanged; 2.0-3.0s linearly rises to 30℃ again; 3.0-4.0s temperature decreases from 30℃ linearly to 25℃ in the afternoon; 4.0-5.0s temperature remains unchanged.

[0165] Figure 5 The simulation results of the present application and Figure 6 compared with the traditional incremental conductance method, the present application has faster tracking speed under the condition of light mutation, the power oscillation amplitude is significantly reduced under the condition of steady state, and the overall energy capture efficiency is effectively improved.

[0166] The above description is only the preferred embodiment of the present application, but the protection scope of the present application is not limited to this. Any skilled person in the art, according to the technical solution and the inventive concept of the present application, makes equivalent replacement or change within the technical range disclosed by the present application, which should be covered in the protection scope of the present application.

Claims

1. A control method for a photovoltaic MPPT dual-mode control system, the photovoltaic MPPT dual-mode control system comprising a photovoltaic module unit, a Boost circuit, a voltage sampling circuit, a current sampling circuit, an MPPT controller, a PWM pulse width modulation module, and a DC electronic load; the photovoltaic module unit is connected to the Boost circuit, the voltage sampling circuit and the current sampling circuit sample the voltage signal and current signal generated by the photovoltaic module unit respectively and send them to the MPPT controller, the MPPT controller uses a dual-mode power prediction control method combining an exponentially adjustable step size conductance increment method and a constant voltage tracking method to automatically adjust the duty cycle, and then sends the optimal duty cycle signal to the PWM pulse width modulation module, and finally outputs a PWM wave through the PWM pulse width modulation module to the power switching transistor IGBT in the Boost circuit, thereby stabilizing the output power of the photovoltaic module unit at its maximum value; characterized in that, The control method includes the following steps: Step 1: Periodically collect the output voltage and output current of the photovoltaic module unit and calculate the output power; Step 2: Construct a disturbance trend prediction quantity using the first-order and second-order power changes. This is used to identify environmental change trends in advance. Step 3: Dynamically adjust the hysteresis interval width of mode switching based on the output power fluctuation variance to achieve noise-adaptive mode determination, based on the disturbance trend prediction. Based on the noise adaptive hysteresis judgment result, select either the incremental conductance mode or the constant voltage tracking mode; Step 4: In constant voltage tracking mode, set the reference voltage based on the open-circuit voltage ratio and adjust the duty cycle through a PI controller; Step 5: In incremental conductance mode, generate an adaptive step size based on the nonlinear potential function and update the duty cycle; Step 6: Dual-mode collaborative mechanism: Combining the improved incremental conductance mode in Step 5 with the constant voltage tracking mode in Step 4, the intelligent selection of the mode is achieved by using the disturbance trend prediction and noise adaptive hysteresis judgment results, so that the system can automatically maintain the optimal control mode under different operating conditions. Step 7: Generate a PWM wave based on the final duty cycle to control the on / off state of the power switch in the Boost circuit, thereby achieving maximum power point tracking.

2. The control method according to claim 1, characterized in that, The photovoltaic module unit includes multiple photovoltaic cells. The Boost circuit includes capacitor C1, capacitor C2, inductor L, diode D, and power switch IGBT. After the photovoltaic module unit is connected to the Boost circuit, positive and negative voltages are generated across capacitor C1. The positive terminal of the photovoltaic module unit is connected to inductor L. The other end of inductor L is connected to the collector of the power switch IGBT and the forward conduction terminal of diode D. The cathode of diode D is connected to one end of capacitor C2 and a DC electronic load. The gate of the power switch IGBT is connected to the output terminal of the PWM pulse width modulation module. The emitter of the power switch IGBT is connected to the negative terminal of the photovoltaic module unit.

3. The control method according to claim 1, characterized in that, Step 1 includes calculating the voltage and current changes at adjacent sampling times, and calculating the power change rate and derivative ratio based on the voltage and current changes, which are used to determine the position of the operating point relative to the maximum power point (MPP).

4. The control method according to claim 1, characterized in that, The predicted disturbance trend in step 2 It is obtained by weighted summation of the first-order and second-order differences of power, and the weighting coefficients are determined experimentally; when When, the system switches to incremental conductance mode; when At this time, the system switches to constant voltage tracking mode; in It is the prediction threshold, which is determined by the width of the adaptive hysteresis interval.

5. The control method according to claim 1, characterized in that, Step 3 first calculates the noise estimate of the output signal, and then dynamically adjusts the hysteresis interval width based on the noise estimate. The greater the noise, the wider the hysteresis interval, and the smaller the noise, the narrower the hysteresis interval.

6. The control method according to claim 1, characterized in that, The parameters of the PI controller described in step 4 are tuned based on a linearized model of the system near the maximum power point, and the proportional gain and integral gain are determined using the step response method.

7. The control method according to claim 1, characterized in that, Step 5 first proposes a nonlinear potential function containing a first-order slope and a second-order curvature based on the geometric characteristics of the photovoltaic power-voltage curve; then, based on the nonlinear potential function, an exponential adaptive step size is designed so that the step size increases rapidly when it is far from the maximum power point (MPP) and automatically shrinks when it is close to the maximum power point (MPP).

8. The control method according to claim 7, characterized in that, The nonlinear potential function is composed of a weighted combination of the absolute values ​​of the first and second derivatives of power with respect to voltage, wherein the second derivative term is used to suppress step overshoot in regions with large curvature.

9. The control method according to claim 7, characterized in that, The exponential adaptive step size performs an exponential mapping between the minimum and maximum step sizes based on the value of the nonlinear potential function, and performs smoothing and limiting processing before output.

Images