Buck-Boost converter circuit for photovoltaic maximum power point tracking

Abstract

The invention discloses a four-switch Buck-Boost converter circuit for photovoltaic maximum power point tracking. The four-switch Buck-Boost converter circuit comprises a main circuit module, an input voltage detection module, an output voltage detection module, an output current detection module, a control output module and a driving amplification circuit module. The control output module comprises a micro-control unit, and the micro-control unit is provided with an ADC sampling interface and a PWM output interface. The main circuit module processes the input voltage of the photovoltaic panel and outputs direct current to charge the battery; the input voltage detection module, the output voltage detection module and the output current detection module are used for sampling the input voltage, the output voltage and the output current of the main circuit module and transmitting sampling values to the micro-control unit; the micro-control unit judges the mode of the main circuit module according to a sampling value, a PWM signal is output by the PWM output interface through an improved maximum power point tracking algorithm, and the on and off of a switching tube of the main circuit module are controlled through the driving amplification circuit module.

Description

Technical Field

[0001] The present invention relates to the field of photovoltaics and switching power supplies, and in particular to a three-mode four-switch Buck-Boost converter circuit for photovoltaic maximum power point tracking. Background Art

[0002] Traditional photovoltaic input circuits are mainly Buck circuits, Boost circuits or Buck-Boost circuits. This puts forward requirements on the input voltage of photovoltaic panels. However, in fact, due to changes in light intensity, temperature, humidity, etc., the input voltage of photovoltaic panels fluctuates within a certain range. Buck circuits, Boost circuits or Buck-Boost circuits cannot fully meet the voltage range of photovoltaic panel inputs. Therefore, a Buck-Boost converter circuit with a wide voltage input range that can be used for photovoltaic maximum power point tracking is proposed.

[0003] The traditional four-switch Buck-Boost converter (FSBB) circuit has Buck and Boost modes, and can switch freely between the two modes. In addition, the use of MOSFET tubes instead of diodes reduces circuit losses and improves efficiency. However, the switch from Buck mode to Boost mode is completed instantly, which will cause irreversible losses to the switch tube devices. Therefore, the Buck-Boost transition mode is added here to smooth the duty cycle switching process, and the method of improving the switching duty cycle is used to further improve the circuit conversion efficiency.

[0004] The maximum power point tracking algorithm is one of the important factors affecting the photovoltaic conversion efficiency. The use of advanced maximum power point tracking (MPPT) algorithm can greatly improve the photovoltaic conversion efficiency. The traditional maximum power point tracking algorithm is mainly based on the conductance increment method and the perturbation observation method. However, the above methods have the disadvantages of empirical trial step setting, poor stability, low output efficiency, low applicability, and easy loss of the maximum power point under shading conditions. Therefore, a new improved intelligent maximum power point tracking algorithm is proposed and innovatively applied to the four-switch Buck-Boost converter circuit. Summary of the invention

[0005] The purpose of the present invention is to overcome the shortcomings and deficiencies of the prior art and provide a Buck-Boost converter circuit for photovoltaic maximum power point tracking, which can significantly increase the input voltage range of the photovoltaic panel and reduce power loss. At the same time, an improved duty cycle switching method and an improved maximum power point tracking algorithm of the transition mode Buck-Boost are adopted to improve the adaptability, stability, reliability, accuracy and efficiency of the circuit, so as to meet the increasing performance requirements of the photovoltaic conversion circuit.

[0006] In order to achieve the above object, the technical solution adopted by the present invention is as follows:

[0007] A Buck-Boost converter circuit for photovoltaic maximum power point tracking includes a main circuit module, an input voltage detection module, an output voltage detection module, an output current detection module, a control output module and a drive amplifier circuit module. The control output module includes a micro control unit, and the micro control unit has an ADC sampling interface and a PWM output interface.

[0008] The input ends of the main circuit module and the input voltage detection module are respectively connected to the photovoltaic panel, the output end of the main circuit module is respectively connected to the input end of the output voltage detection module, the output current detection module and the battery, and the output ends of the input voltage detection module, the output voltage detection module and the output current detection module are respectively connected to the ADC sampling interface;

[0009] The main circuit module is used to process the 15-130V input voltage of the photovoltaic panel and output direct current to charge the battery; the input voltage detection module, the output voltage detection module, and the output current detection module are used to sample the input voltage, output voltage, and output current of the main circuit module, and transmit the sampled values ​​to the microcontroller unit; the microcontroller unit is used to determine the mode of the main circuit module according to the sampled values, and output the PWM control signal from the PWM output interface by improving the maximum power point tracking algorithm, and control the on and off of the switch tube of the main circuit module through the driving amplifier circuit module;

[0010] The main circuit module includes a first switch tube, a second switch tube, a third switch tube, a fourth switch tube, a first diode, a second diode, a third diode, a fourth diode, a first parasitic capacitor, a second parasitic capacitor, a third parasitic capacitor, a fourth parasitic capacitor and a charge and discharge inductor;

[0011] A first PWM control signal is input to the control end of the first switch tube, a first end of the first switch tube is connected to the positive electrode of the first diode and the first parasitic capacitor, and a second end of the first switch tube is connected to the negative electrode of the first diode and the first parasitic capacitor;

[0012] The control end of the second switch tube inputs the second PWM control signal, the first end of the second switch tube is connected to the second diode and the positive electrode of the second parasitic capacitor, and the second end of the second switch tube is connected to the second diode and the negative electrode of the second parasitic capacitor;

[0013] A third PWM control signal is input to the control end of the third switch tube, a first end of the third switch tube is connected to the positive electrode of the third diode and the third parasitic capacitor, and a second end of the third switch tube is connected to the negative electrode of the third diode and the third parasitic capacitor;

[0014] The control end of the fourth switch tube inputs the fourth PWM control signal, the first end of the fourth switch tube is connected to the positive electrode of the fourth diode and the fourth parasitic capacitor, and the second end of the fourth switch tube is connected to the negative electrode of the fourth diode and the fourth parasitic capacitor.

[0015] Furthermore, the input voltage detection module uses the operational amplifier TP2412 for differential sampling to reduce the input voltage by 50 times. When the maximum input voltage is 130V, the maximum output voltage is 2.6V after the differential sampling circuit.

[0016] Furthermore, the output voltage detection module uses the operational amplifier TP2412 for differential sampling to reduce the input voltage by 20 times. When the maximum output voltage is 55V, the maximum output voltage is 2.75V after the differential sampling circuit.

[0017] Furthermore, the output current detection module connects two 2mR / 2W current sensing resistors in parallel, and the current is shunted through the two resistors. The maximum current at the output end is 63A and the power reaches 3000W. Through the INA280 current detection amplifier, the voltage drop on the shunt resistor is measured in a wide common mode range of 2.7V to 120V, and the sampled value is input into the microcontroller unit.

[0018] Furthermore, the Buck-Boost converter circuit has three operating modes, specifically:

[0019] Let V i is the input voltage, V o is the nominal voltage at the output terminal. i Less than V o -ΔV h When V i Greater than V o +ΔVh, the circuit is in Buck mode; when V i At V o -ΔV h and V o +ΔV h When ΔV h It is half of the output voltage range of Buck-Boost mode;

[0020] The Buck-Boost operating mode of the Buck-Boost converter circuit is between the Buck and Boost modes. When the circuit transitions from the Buck mode to the Boost mode, the intermediate process is the Buck-Boost operating mode.

[0021] Buck-Boost mode range depends on the voltage gain M, expressed as:

[0022]

[0023] Wherein, D1 is the duty cycle of the first switch tube; D2 is the duty cycle of the fourth switch tube;

[0024] The relationship between D1 and D2 is expressed as:

[0025]

[0026] Let D1 and D2 satisfy the linear relationship:

[0027] D1=D2+0.95;

[0028] This eliminates the jump process of the operating point, and the voltage gain M is:

[0029]

[0030] Furthermore, the optimized particle swarm algorithm BSPSO is used to track the maximum power point under partial shadow conditions, specifically:

[0031] First, chaos initialization based on the reverse strategy is carried out, and the improved Tent mapping and quasi-reverse individuals are used for initialization; then the evolutionary mutation stage is carried out, and the mutation factor is introduced to further expand the range of the global optimal solution; finally, the optimization convergence stage is entered, and the speed term is removed to enter the low-latitude optimization, which accelerates the convergence speed until it converges to the global optimal solution; the voltage and current output by the photovoltaic panel are taken as independent variables, the power is taken as the objective function, and the output duty cycle is used to control the maximum output power.

[0032] Furthermore, the chaotic initialization process based on the reverse strategy of the particle swarm optimization BSPSO is optimized, including:

[0033] The particle swarm PSO algorithm is a global optimization algorithm based on swarm intelligence. The update formula during the particle working process is:

[0034] v n i =wv n-1 i +c1r1(p n i -x ni )+c2r2(g n -x n i );

[0035] x n+1 i =x n i +v n i ;

[0036] In the formula, v n i is the velocity of particle i at the nth iteration; w is the inertia weight; v n-1 i is the speed of particle i at the n-1th iteration; c1 is the individual learning factor; r1 and r2 are random numbers between 0 and 1; p n i is the optimal position of particle i after n iterations; c2 is the group learning factor; g n is the best position of all particles after n iterations; x n i is the position of particle i at the nth iteration; x n+1 i is the position of particle i at the n+1th iteration;

[0037] Using improved Tent mapping to initialize particle swarm chaos:

[0038]

[0039] In the formula, x i is the initial position of particle i; x i ～ is the position of particle i after Tent mapping;

[0040] The condition for this chaotic mapping to be valid is that it is a full mapping in [0,1], which is related to s and q, where the value of q is a discrete point set:

[0041] q=2πN N∈(-∞,0)∪(0,+∞),N∈Z;

[0042] The value of s:

[0043]

[0044] The quasi-reverse strategy is used to initialize the individual particle swarms. The chaos initialization based on the reverse strategy is:

[0045]

[0046] In the formula, x i^ is the position of particle i after mapping; x mid is the center of the feasible region, rand represents the random function of (0,1), and sin represents the sin function.

[0047] Furthermore, the evolution and mutation phase of the particle swarm optimization BSPSO is optimized, including:

[0048] Although the use of chaotic initialization particle swarm based on the reverse strategy can speed up the convergence speed, the similarity of the particle swarm increases during the optimization process, and it is easy to fall into local optimization. In order to jump out of the local optimization and improve the diversity of the particle swarm, the optimization idea of ​​the Tianniu algorithm is used for improvement;

[0049] Without knowing the exact location of the food, the longhorn beetle senses it through the antennae on both sides of its head. Since the antennae are at different distances from the food, the concentration of the odor sensed is also different. The longhorn beetle can move towards the food based on the concentration difference of the sensed odor and reach the food after multiple iterations. The mathematical expression is:

[0050]

[0051] In the formula, is a random unit vector, indicating the direction of the beetle's whiskers at any time; rands represents a random vector of (-1,1); Dim represents the spatial dimension; X t+1 represents the centroid position of the beetle after n iterations at time t+1; X l , X r X is the position of the left and right antennae of the longicorn beetle; t represents the centroid position of the beetle after n iterations at time t; f function is the objective function; δt is the iteration step length; sign is the sign function;

[0052] In order to improve the global optimization of particle swarm, the particle swarm algorithm formula is improved:

[0053] v n i =wv n-1 i +c1r1(p n i -x n i )+c2r2(g n -x n i )+c3r3(p n i -g n );

[0054] In the formula, c3 is the variation factor; r3 is a random number between 0 and 1;

[0055] The optimal positions of the current particle and the group are introduced in the formula, the positions of the two whiskers are regarded as the local optimum of the individual and the local optimum of the group, and the distance between the two whiskers represents the distance between the optimal positions of the individual and the group.

[0056] Furthermore, the optimization convergence phase of the particle swarm optimization algorithm BSPSO is optimized, including:

[0057] After iteration, when the best position found by an individual particle swarm is better than the best position obtained by the group in the previous iteration, the particle swarm in the vicinity of the individual enters the optimization convergence state, switches the iterative algorithm, uses the simplified particle swarm PSO algorithm, removes the speed term, and speeds up the convergence speed;

[0058] If we only consider the motion of the i-th particle, we can update the formula for the particle swarm by

[0059]

[0060] It can be changed to:

[0061]

[0062] x(n+1)=x(n)+v(n);

[0063] After iterating the above formula, we get:

[0064]

[0065] The formula is a classical second-order differential equation without velocity terms;

[0066] At this time, it can be transformed into a first-order differential equation:

[0067]

[0068] The particle swarm optimization equation without velocity term is simplified to:

[0069] x n i =wx n-1 i +c1r1(p n i -x n i )+c2r2(g n -x n i );

[0070] After removing the velocity term, the particle swarm equation is reduced from a second-order differential equation to a first-order differential equation, which simplifies the iterative process and facilitates the analysis and control of the particle swarm evolution process.

[0071] Furthermore, the adaptive parameter process of the particle swarm algorithm BSPSO is optimized, including:

[0072] Adopt nonlinear strategy to update inertia weight w:

[0073]

[0074] Where w1 is the maximum value of the inertia weight w; w2 is the minimum value of the inertia weight w; h is the nonlinear coefficient; k m is the maximum number of iterations set; k is the current number of iterations;

[0075] The individual learning factor c1 and the group learning factor c2 are updated using the trigonometric formula:

[0076]

[0077] The mutation factor c3 is updated using a nonlinear strategy:

[0078]

[0079] In the formula, c α =0.9, c β =0.45.

[0080] Compared with the prior art, the present invention has the following advantages and beneficial effects.

[0081] (1) The present invention increases the input voltage range of the photovoltaic panel by adopting a four-switch Buck-Boost converter circuit to replace the traditional Buck circuit, Boost circuit or Buck-Boost circuit, thereby expanding the input voltage range of the photovoltaic panel to 15-130V.

[0082] (2) The present invention adopts MOSFET switch tube to replace the traditional diode, and controls the on and off of the MOSFET switch tube through PWM. There is no conduction voltage drop and conduction loss of the diode, which improves the circuit power conversion efficiency.

[0083] (3) The microcontroller unit of the present invention adopts a dsPIC chip with higher precision and reliability, has high performance and flexible peripherals, and exhibits powerful performance in asynchronous event processing capability, precision simulation, etc., thereby improving the precision and reliability of the circuit.

[0084] (4) The present invention adds a transition mode, Buck-Boost mode, between Buck and Boost modes, which smoothes the transition process, switches the duty cycle smoothly without jumps, reduces device losses, and improves circuit stability.

[0085] (5) The present invention adopts an intelligent improved particle swarm algorithm to replace the conductance increment method and the perturbation observation method, so as to further improve the maximum power point tracking effect of the photovoltaic panel under the shade state, achieve higher output power, and improve the maximum power point tracking effect of the circuit. BRIEF DESCRIPTION OF THE DRAWINGS

[0086] Figure 1 Schematic diagram of the structure of the Buck-Boost converter circuit used for photovoltaic maximum power point tracking.

[0087] Figure 2 This is the sampling circuit diagram of the input voltage detection module.

[0088] Figure 3 This is the sampling circuit diagram of the output voltage detection module.

[0089] Figure 4 This is the sampling circuit diagram of the output current detection module.

[0090] Figure 5 Schematic diagram of the three modes of the four-switch Buck-Boost converter circuit.

[0091] Figure 6 Schematic diagram of duty cycle switching of three modes of a four-switch Buck-Boost converter circuit.

[0092] Figure 7 Schematic diagram of the Buck-Boost mode duty cycle switching continuous strategy for a four-switch Buck-Boost converter circuit.

[0093] Figure 8 Schematic diagram of the PV curve of photovoltaic cells under shade conditions.

[0094] Fig. 9 Flowchart of particle swarm optimization algorithm.

[0095] Fig.10 Schematic diagram of the simulation for optimizing the particle swarm algorithm.

[0096] Fig.11 Figure 2. Simulation experiment results to verify the optimization of particle swarm algorithm. DETAILED DESCRIPTION

[0097] The Buck-Boost converter circuit for photovoltaic maximum power point tracking of the present invention is further described below in conjunction with the accompanying drawings and specific embodiments.

[0098] See also Figure 1The present invention discloses a Buck-Boost converter circuit for photovoltaic maximum power point tracking, including a main circuit module, an input voltage detection module, an output voltage detection module, an output current detection module, a control output module and a drive amplifier circuit module. The control output module includes a micro control unit, and the micro control unit has an ADC sampling interface and a PWM output interface.

[0099] The microcontroller uses the dsPIC33CK128MP205 chip. The dsPIC is a device that combines the features of a single-chip microcomputer with the capabilities of a digital signal processor (DSP). It has high performance, flexible peripherals, and a complete ecosystem of software and hardware tools. It has strong performance in asynchronous event processing capabilities, precision simulation, common development environments, and peripheral components.

[0100] The input ends of the main circuit module and the input voltage detection module are respectively connected to the photovoltaic panels, the output ends of the main circuit module are respectively connected to the input ends of the output voltage detection module, the output current detection module and the battery, and the output ends of the input voltage detection module, the output voltage detection module and the output current detection module are respectively connected to the ADC sampling interface.

[0101] The main circuit module is used to process the 15-130V input voltage of the photovoltaic panel and output direct current to charge the battery; the input voltage detection module, output voltage detection module, and output current detection module are used to sample the input voltage, output voltage, and output current of the main circuit module, and transmit the sampled values ​​to the microcontroller unit; the microcontroller unit is used to determine the mode of the main circuit module according to the sampled values, and output the PWM control signal from the PWM output interface by improving the maximum power point tracking algorithm, and control the on and off of the switch tube of the main circuit module through the driving amplifier circuit module.

[0102] The main circuit module includes a first switch tube Q1, a second switch tube Q2, a third switch tube Q3, a fourth switch tube Q4, a first diode, a second diode, a third diode, a fourth diode, a first parasitic capacitor, a second parasitic capacitor, a third parasitic capacitor, a fourth parasitic capacitor and a charge and discharge inductor L.

[0103] The control end of the first switch tube Q1 inputs the first PWM control signal PWM1, the first end of the first switch tube Q1 is connected to the positive electrode of the first diode and the first parasitic capacitor, and the second end of the first switch tube Q1 is connected to the negative electrode of the first diode and the first parasitic capacitor.

[0104] The control end of the second switch tube Q2 inputs the second PWM control signal PWM2, the first end of the second switch tube Q2 is connected to the second diode and the positive electrode of the second parasitic capacitor, and the second end of the second switch tube Q2 is connected to the second diode and the negative electrode of the second parasitic capacitor.

[0105] The control end of the third switch tube Q3 inputs the third PWM control signal PWM3, the first end of the third switch tube Q3 is connected to the positive electrode of the third diode and the third parasitic capacitor, and the second end of the third switch tube Q3 is connected to the negative electrode of the third diode and the third parasitic capacitor.

[0106] The control end of the fourth switch tube Q4 inputs the fourth PWM control signal PWM4, the first end of the fourth switch tube Q4 is connected to the anode of the fourth diode and the fourth parasitic capacitor, and the second end of the fourth switch tube Q4 is connected to the cathode of the fourth diode and the fourth parasitic capacitor.

[0107] like Figure 2 As shown in the figure, the input voltage detection module uses the operational amplifier TP2412 for differential sampling, which reduces the input voltage by 50 times. When the maximum input voltage is 130V, the maximum output voltage is 2.6V after the differential sampling circuit. At the same time, in order to protect the microcontroller unit, a diode is added to the ADC sampling interface of the microcontroller unit for clamping protection, so that the voltage of the input ADC sampling interface does not exceed 3.3V.

[0108] like Figure 3 As shown in the figure, the output voltage detection module uses the operational amplifier TP2412 for differential sampling, which reduces the input voltage by 20 times. When the maximum output voltage is 55V, the maximum output voltage is 2.75V after the differential sampling circuit. At the same time, in order to protect the microcontroller unit, a diode is added to the ADC sampling interface of the microcontroller unit for clamping protection, so that the voltage of the input ADC sampling interface does not exceed 3.3V.

[0109] like Figure 4 As shown in the figure, the output current detection module connects two 2mR / 2W current sensing resistors in parallel. Through the two resistors, the maximum current at the output end is 63A, and the power can reach 3000W. Through the INA280 current detection amplifier, the voltage drop on the shunt resistor can be measured in a wide common mode range of 2.7V to 120V, and the sampled value can be input to the microcontroller unit.

[0110] like Figure 5 As shown in the figure, the four-switch Buck-Boost converter circuit presents three operating modes. i is the input voltage, V o is the nominal voltage at the output terminal. i Less than V o -ΔV hWhen V i Greater than V o +ΔVh, the circuit is in Buck mode; when V i At V o -ΔV h and V o +ΔV h When ΔV h Half of the output voltage range of the Buck-Boost mode; usually V o *5% (duty cycle), here we take V o is 50, then V o -ΔV h is 47.5, V o +ΔV h It is 52.5.

[0111] like Figure 6 The figure shows the mode switching diagram of the four-switch Buck-Boost converter circuit. The Buck-Boost operating mode of the four-switch Buck-Boost converter circuit is proposed to solve the voltage fluctuation caused by the frequent switching of the dual-mode Buck and Boost modes. The Buck-Boost operating mode is between the Buck and Boost modes. When the circuit transitions from the Buck mode to the Boost mode, the intermediate process is the Buck-Boost operating mode. The Buck-Boost mode range depends on the voltage gain M, which is expressed as:

[0112]

[0113] When the input voltage V i When the input voltage V i When the output voltage is greater than 52.5V and the duty cycle D2 of the fourth switch tube Q4 is greater than 5%, the circuit is in Boost mode, the second switch tube Q2 is not turned on, D1 is 1, and the voltage gain M is 1 / (1-D2). When the output voltage range is (47.5V, 52.5V), the circuit is in Buck-Boost mode. At this time, in one cycle, the first switch tube Q1 and the fourth switch tube Q4 are turned on first, the inductor stores energy, and the inductor current rises. Then, the second switch tube Q2 and the third switch tube Q3 are turned on again, the inductor L discharges, and the inductor L current decreases.

[0114] In the ideal mode, the influence of dead zone is not considered. Since the input voltage is not sampled, the startup process of the four-switch Buck-Boost converter circuit is first Buck mode, then switch to Buck-Boost mode, and finally to Boost mode. The change of duty cycle D1 and D2 in the intermediate process of Buck-Boost mode has always been the focus of research. The relationship between the two can be expressed as:

[0115]

[0116] It can be seen from the formula that there is a linear relationship between duty cycles D1 and D2. Plotting with duty cycle D1 as the horizontal axis and duty cycle D2 as the vertical axis, as shown in Figure 6 As shown, the variation range of the duty cycles D1 and D2 in the Buck-Boost mode is a quadrilateral area surrounded by two straight lines with slopes of -0.95 and -1.05. It can be seen that the switching modes of the duty cycles D1 and D2 are various, which determines the various strategies for Buck-Boost mode switching.

[0117] like Figure 7 As shown, the present invention proposes a new Buck-Boost transition mode duty cycle switching strategy, eliminating the jump process of the working point. In Buck mode, the duty cycle is switched along the OA segment to point A, at which the duty cycles D1 and D2 are (0.95, 0), and then the circuit is switched to the Buck-Boost mode, and the duty cycle is switched along the AD segment to point D, at which the duty cycles D1 and D2 are (1, 0.05). At this time, the circuit is switched to the Boost mode, and the duty cycle is switched along the DE segment, then each duty cycle is continuous without jump, which ensures the continuity of the duty cycle switching and improves the stability of the circuit.

[0118] Analyzing the AD segment, the duty cycles D1 and D2 of the AD segment satisfy a linear relationship:

[0119] D1=D2+0.95;

[0120] Wherein, D1 is the duty cycle of the first switch tube Q1; D2 is the duty cycle of the fourth switch tube Q4.

[0121] The voltage gain M is:

[0122]

[0123] The duty cycle D1 range is (0.95, 1), and the voltage gain M range is (0.95, 1.05).

[0124] like Figure 8As shown in the figure, under partial shadow conditions, the operating voltage and current of the shaded components are reduced, resulting in multiple local extreme points in the PV curve of the photovoltaic cell. The photovoltaic cell power model is built using three photovoltaic panels in series. When the light intensity of the series photovoltaic panels is 1000W / m2, 800W / m2, and 600W / m2 respectively, and the ambient temperature is 25℃, the PV curve is as follows Figure 8 At this time, the PV curve has multiple peaks, and the conventional optimization algorithm can no longer meet the multi-peak optimization requirements.

[0125] like Fig. 9 As shown, the present invention adopts the optimized particle swarm algorithm BSPSO (Behavior Strategies ParticleSwarm Optimization) to track the maximum power point under local shadow conditions. First, chaos initialization based on the reverse strategy is performed, and the improved Tent mapping and quasi-reverse individuals are used for initialization; then the evolutionary mutation stage is carried out, and the mutation factor is introduced to further expand the range of the global optimal solution; finally, the optimization convergence stage is entered, the speed term is removed to enter the low-latitude optimization, and the convergence speed is accelerated until it converges to the global optimal solution; the voltage and current output by the photovoltaic panel are used as independent variables, the power is used as the objective function, and the output duty cycle is used to control the maximum output power.

[0126] (1) Particle Swarm Optimization

[0127] Particle Swarm Optimization (PSO) algorithm is a global optimization algorithm based on swarm intelligence. The update formula during particle working is:

[0128] v n i =wv n-1 i +c1r1(p n i -x n i )+c2r2(g n -x n i );

[0129] x n+1 i =x n i +v n i ;

[0130] In the formula, v n i is the velocity of particle i at the nth iteration; w is the inertia weight; v n-1 iis the speed of particle i at the n-1th iteration; c1 is the individual learning factor; r1 and r2 are random numbers between 0 and 1; p n i is the optimal position of particle i after n iterations; c2 is the group learning factor; g n is the best position of all particles after n iterations; x n i is the position of particle i at the nth iteration; x n+1 i is the position of particle i at the n+1th iteration.

[0131] (2) Chaos initialization

[0132] Chaos is an unstable phenomenon that occurs spontaneously in a deterministic system. Chaotic sequences are characterized by regularity, randomness, and ergodicity. The PSO population initialization that integrates the chaos algorithm has better diversity and convergence. The chaos algorithm mainly generates new chaotic sequences through nonlinear iteration of chaotic mapping in the interval [0, 1]. There are many types of chaotic mapping, including Logistic mapping, Tent mapping, Circle mapping, Cubic mapping, etc.

[0133] The present invention adopts improved Tent mapping to perform particle swarm chaos initialization:

[0134]

[0135] In the formula, x i is the initial position of particle i; x i ～ is the position of particle i after Tent mapping.

[0136] The condition for this chaotic mapping to be valid is that it is a full mapping in [0,1], which is related to s and q, where the value of q is a discrete point set:

[0137] q=2πn n∈(-∞,0)∪(0,+∞),n∈Z;

[0138] The value of s:

[0139]

[0140] (3) Reverse strategy

[0141] According to the principle of probability, each randomly generated candidate solution has a 50% probability of being far away from or close to the optimal solution compared to its reverse solution. The reverse solution may also be the optimal solution. Selecting individuals that are closer to the two as initial population members will accelerate convergence to a certain extent.

[0142] The present invention adopts a quasi-reverse strategy to initialize the particle swarm individuals, and the corresponding standard reverse individual positions are:

[0143] x i * =2x mid -x i ;

[0144] In the formula, x i * is the position of particle i after standard reverse individual mapping; x mid is the center of the feasible region; x i is the initial position of particle i.

[0145] The quasi-reverse individual position is expressed as:

[0146] x i '=rand(x mid ,x i * );

[0147] In the formula, x i ' is the position of particle i after quasi-reverse individual mapping; rand represents the random function of (0, 1).

[0148] The quasi-reflective reverse individual position is expressed as:

[0149] x i = rand(x i ,x mid );

[0150] In the formula, x i ” is the position of particle i after quasi-reflection reverse individual mapping.

[0151] Then the chaos initialization based on the reverse strategy is:

[0152]

[0153] In the formula, x i ^ is the position of particle i after mapping; sin represents the sin function.

[0154] (4) Evolutionary variation

[0155] Although the use of chaotic initialization particle swarm based on the reverse strategy can accelerate the convergence speed to a certain extent, the similarity of particle swarm increases during the optimization process, which makes it easy to fall into local optimization. In order to jump out of local optimization and improve the diversity of particle swarm, the optimization idea of ​​Tianniu algorithm is used for improvement.

[0156] Without knowing the exact location of the food, the longhorn beetle senses it through the antennae on both sides of its head. Since the antennae are at different distances from the food, the concentration of the odor sensed is also different. The longhorn beetle can move towards the food based on the concentration difference of the sensed odor and reach the food after multiple iterations. The mathematical expression is:

[0157]

[0158] In the formula, is a random unit vector, indicating the direction of the beetle's whiskers at any time; rands represents a random vector of (-1, 1); Dim represents the spatial dimension; X t+1 represents the centroid position of the beetle after n iterations at time t+1; X l , X r X is the position of the left and right antennae of the longicorn beetle; t It represents the centroid position of the beetle after n iterations at time t; f function is the objective function; δt is the iteration step size; sign is the sign function.

[0159] In order to improve the global optimization of particle swarm, the particle swarm algorithm formula is improved:

[0160] v n i =wv n-1 i +c1r1(p n i -x n i )+c2r2(g n -x n i )+c3r3(p n i -g n );

[0161] In the formula, c3 is the variation factor; r3 is a random number between 0 and 1;

[0162] The optimal positions of the current particle and the group are introduced in the formula, the positions of the two whiskers are regarded as the local optimum of the individual and the local optimum of the group, and the distance between the two whiskers represents the distance between the optimal positions of the individual and the group.

[0163] (5) Optimizing Convergence

[0164] After iteration, when the best position found by an individual particle swarm is better than the best position obtained by the group in the previous iteration, the particle swarm in the vicinity of the individual enters the optimization convergence state, converts the iterative algorithm, adopts the simplified particle swarm PSO algorithm, removes the speed term, and speeds up the convergence speed.

[0165] If we only consider the motion of the i-th particle, we can update the formula for the particle swarm by

[0166]

[0167] It can be changed to:

[0168]

[0169] x(n+1)=x(n)+v(n);

[0170] After iterating the above formula, we get:

[0171]

[0172] The formula is a classical second-order differential equation without velocity terms.

[0173] At this time, it transforms into a first-order differential equation:

[0174]

[0175] The particle swarm optimization equation without velocity term is simplified to:

[0176] x i n =wx i n-1 +c1r1(p i n -x i n )+c2r2(g n -x i n );

[0177] It can be seen that after removing the velocity term, the particle swarm equation is reduced from a second-order differential equation to a first-order differential equation, which simplifies the iterative process and facilitates the analysis and control of the particle swarm evolution process. The convergence accuracy and convergence speed of the simplified particle swarm algorithm (SPSO) have been significantly improved, so the use of the SPSO algorithm in the convergence stage of particle swarm optimization can significantly improve the convergence speed.

[0178] (6) Adaptive parameters

[0179] The values ​​of the three important parameters in the particle swarm algorithm, inertia weight w, individual learning factor c1 and group learning factor c2, have a great influence on the maximum power point tracking process.

[0180] When the inertia weight w is small, the local optimization ability of the particles is improved, which accelerates the convergence of the algorithm; when the inertia weight w is large, the global optimization ability is improved. For individual learning factor c1 and group learning factor c2, individual learning is more important in the initial stage of the algorithm, while group learning is more important in the later stage.

[0181] The present invention adopts a nonlinear strategy to update the inertia weight w:

[0182]

[0183] Where w1 is the maximum value of the inertia weight w; w2 is the minimum value of the inertia weight w; t is the nonlinear coefficient; k m is the maximum number of iterations set; k is the current number of iterations.

[0184] The individual learning factor c1 and the group learning factor c2 are updated using the trigonometric formula:

[0185]

[0186] The mutation factor c3 is updated using a nonlinear strategy:

[0187]

[0188] In the formula, c α =0.9, c β =0.45.

[0189] The optimization algorithm flow chart is as follows Fig. 9 shown.

[0190] like Fig.10 As shown in the figure, when the local shadow condition occurs, the light intensity suddenly changes, and the PV curve has multiple local extreme points. In order to verify the global optimization of the BSPSO algorithm, a local shadow disturbance is performed at 0.5s to change the light intensity of the third photovoltaic panel from 600W / m 2 Down to 400W / m 2 , to test the algorithm’s ability to escape from local extreme values ​​and perform global optimization.

[0191] The initial stabilization time of the BSPSO algorithm is 0.032s, the initial stable power is 4255W, the second stable time is 0.517s, the second stable power is 2208W, and the power fluctuation is 787W. The BSPSO algorithm is 72% faster than the PSO algorithm in tracking the maximum power point. At the same time, the BSPSO algorithm can jump out of the local optimal point. When the disturbance is applied, the BSPSO algorithm recovers the stable state time and the fluctuation amplitude are reduced, and the algorithm is significantly improved.

[0192] Connect the programmable power supply, four-switch Buck-Boost converter circuit board, battery, and load meter. Set the maximum power point to 77V, 14.3A, and 1100W. Simulate the shadow state at 16s, and the light intensity drops by half. It returns to normal after 26s. Then test the algorithm to find the best solution. Fig.11As shown, the real-time tracking status of voltage, current, and power, as well as the maximum power point tracking efficiency, it can be seen that the minimum efficiency is above 90%. Due to the sudden drop in light intensity, the efficiency decreases. The efficiency in other time periods is above 98%, and the average efficiency is above 96%. The algorithm has good optimization performance.

[0193] In summary, the present invention has the following advantages and beneficial effects.

[0194] (1) The present invention increases the input voltage range of the photovoltaic panel by adopting a four-switch Buck-Boost converter circuit to replace the traditional Buck circuit, Boost circuit or Buck-Boost circuit, thereby expanding the input voltage range of the photovoltaic panel to 15-130V.

[0195] (2) The present invention adopts MOSFET switch tube to replace the traditional diode, and controls the on and off of the MOSFET switch tube through PWM. There is no conduction voltage drop and conduction loss of the diode, which improves the circuit power conversion efficiency.

[0196] (3) The microcontroller unit of the present invention adopts a dsPIC chip with higher precision and reliability, has high performance and flexible peripherals, and exhibits powerful performance in asynchronous event processing capability, precision simulation, etc., thereby improving the precision and reliability of the circuit.

[0197] (4) The present invention adds a transition mode, Buck-Boost mode, between Buck and Boost modes, which smoothes the transition process, switches the duty cycle smoothly without jumps, reduces device losses, and improves circuit stability.

[0198] (5) The present invention adopts an intelligent improved particle swarm algorithm to replace the conductance increment method and the perturbation observation method, so as to further improve the maximum power point tracking effect of the photovoltaic panel under the shade state, achieve higher output power, and improve the maximum power point tracking effect of the circuit.

[0199] The above description is a detailed description of the preferred feasible embodiments of the present invention, but the embodiments are not intended to limit the scope of the patent application of the present invention. All equivalent changes or modified changes completed under the technical spirit disclosed by the present invention should fall within the patent scope covered by the present invention.

Claims

1. A Buck-Boost converter circuit for photovoltaic maximum power point tracking, characterized in that: It includes a main circuit module, an input voltage detection module, an output voltage detection module, an output current detection module, a control output module and a drive amplifier circuit module. The control output module includes a micro control unit, and the micro control unit has an ADC sampling interface and a PWM output interface; The input ends of the main circuit module and the input voltage detection module are respectively connected to the photovoltaic panel, the output end of the main circuit module is respectively connected to the input end of the output voltage detection module, the output current detection module and the battery, and the output ends of the input voltage detection module, the output voltage detection module and the output current detection module are respectively connected to the ADC sampling interface; The main circuit module is used to process the 15-130V input voltage of the photovoltaic panel and output direct current to charge the battery; the input voltage detection module, the output voltage detection module, and the output current detection module are used to sample the input voltage, output voltage, and output current of the main circuit module, and transmit the sampled values ​​to the microcontroller unit; the microcontroller unit is used to determine the mode of the main circuit module according to the sampled values, and output the PWM control signal from the PWM output interface by improving the maximum power point tracking algorithm, and control the on and off of the switch tube of the main circuit module through the driving amplifier circuit module; The main circuit module includes a first switch tube, a second switch tube, a third switch tube, a fourth switch tube, a first diode, a second diode, a third diode, a fourth diode, a first parasitic capacitor, a second parasitic capacitor, a third parasitic capacitor, a fourth parasitic capacitor and a charge and discharge inductor; A first PWM control signal is input to the control end of the first switch tube, a first end of the first switch tube is connected to the positive electrode of the first diode and the first parasitic capacitor, and a second end of the first switch tube is connected to the negative electrode of the first diode and the first parasitic capacitor; The control end of the second switch tube inputs the second PWM control signal, the first end of the second switch tube is connected to the second diode and the positive electrode of the second parasitic capacitor, and the second end of the second switch tube is connected to the second diode and the negative electrode of the second parasitic capacitor; A third PWM control signal is input to the control end of the third switch tube, a first end of the third switch tube is connected to the positive electrode of the third diode and the third parasitic capacitor, and a second end of the third switch tube is connected to the negative electrode of the third diode and the third parasitic capacitor; The control end of the fourth switch tube inputs the fourth PWM control signal, the first end of the fourth switch tube is connected to the positive electrode of the fourth diode and the fourth parasitic capacitor, and the second end of the fourth switch tube is connected to the negative electrode of the fourth diode and the fourth parasitic capacitor.

2. The Buck-Boost converter circuit for photovoltaic maximum power point tracking according to claim 1, characterized in that: The input voltage detection module uses the operational amplifier TP2412 for differential sampling, which reduces the input voltage by 50 times. When the maximum input voltage is 130V, the maximum output voltage is 2.6V after the differential sampling circuit.

3. The Buck-Boost converter circuit for photovoltaic maximum power point tracking according to claim 1, characterized in that: The output voltage detection module uses the operational amplifier TP2412 for differential sampling, which reduces the input voltage by 20 times. When the maximum output voltage is 55V, the maximum output voltage is 2.75V after the differential sampling circuit.

4. The Buck-Boost converter circuit for photovoltaic maximum power point tracking according to claim 1, characterized in that: The output current detection module connects two 2mR / 2W current-sense resistors in parallel. The current is shunted by the two resistors. The maximum current at the output end is 63A and the power reaches 3000W. The voltage drop on the shunt resistor is measured in a wide common-mode range of 2.7V to 120V through the INA280 current detection amplifier, and the sampled value is input into the microcontroller unit.

5. The Buck-Boost converter circuit for photovoltaic maximum power point tracking according to claim 1, characterized in that: Buck-Boost converter circuit has three working modes, specifically: Let V i is the input voltage, V o is the nominal voltage at the output terminal. i Less than V o -ΔV h When V i Greater than V o +ΔVh, the circuit is in Buck mode; when V i At V o -ΔV h and V o +ΔV h When ΔV h It is half of the output voltage range of Buck-Boost mode; The Buck-Boost operating mode of the Buck-Boost converter circuit is between the Buck and Boost modes. When the circuit transitions from the Buck mode to the Boost mode, the intermediate process is the Buck-Boost operating mode. Buck-Boost mode range depends on the voltage gain M, which is expressed as: Wherein, D1 is the duty cycle of the first switch tube; D2 is the duty cycle of the fourth switch tube; The relationship between D1 and D2 is expressed as: Let D1 and D2 satisfy the linear relationship: D1=D2+0.95; This eliminates the jump process of the operating point, and the voltage gain M is:

6. The Buck-Boost converter circuit for photovoltaic maximum power point tracking according to claim 1, characterized in that: The optimized particle swarm algorithm BSPSO is used to track the maximum power point under partial shadow conditions, specifically: Firstly, chaos initialization based on the reverse strategy is carried out, and the improved Tent mapping and quasi-reverse individuals are used for initialization; then the evolutionary mutation stage is carried out, and the mutation factor is introduced to further expand the range of the global optimal solution; finally, the optimization convergence stage is entered, and the speed term is removed to enter the low-latitude optimization, which accelerates the convergence speed until it converges to the global optimal solution; the voltage and current output by the photovoltaic panel are taken as independent variables, the power is taken as the objective function, and the output PWM duty cycle is used to control the maximum output power.

7. The Buck-Boost converter circuit for photovoltaic maximum power point tracking according to claim 6, characterized in that: The chaotic initialization process based on the reverse strategy of the optimized particle swarm algorithm BSPSO includes: The particle swarm PSO algorithm is a global optimization algorithm based on swarm intelligence. The update formula during the particle working process is: v n i =wv n-1 i +c1r1(p n i -x n i )+c2r2(g n -x n i ); x n+1 i =x n i +v n i ; In the formula, v n i is the velocity of particle i at the nth iteration; w is the inertia weight; v n-1 i is the speed of particle i at the n-1th iteration; c1 is the individual learning factor; r1 and r2 are random numbers between 0 and 1; p n i is the optimal position of particle i after n iterations; c2 is the group learning factor; g n is the best position of all particles after n iterations; x n i is the position of particle i at the nth iteration; x n+1 i is the position of particle i at the n+1th iteration; Using improved Tent mapping to initialize particle swarm chaos: In the formula, x i is the initial position of particle i; x i ～ is the position of particle i after Tent mapping; The condition for this chaotic mapping to be valid is that it is a full mapping in [0,1], which is related to s and q, where the value of q is a discrete point set: q=2πN N∈(-∞,0)∪(0,+∞),N∈Z; The value of s: The quasi-reverse strategy is used to initialize the individual particle swarms. The chaos initialization based on the reverse strategy is: In the formula, x i ^ is the position of particle i after mapping; x mid is the center of the feasible region, rand represents the random function of (0,1), and sin represents the sin function.

8. The Buck-Boost converter circuit for photovoltaic maximum power point tracking according to claim 7, characterized in that: The evolution and mutation phase of the optimized particle swarm algorithm BSPSO includes: Although the use of chaotic initialization particle swarm based on the reverse strategy can speed up the convergence speed, the similarity of the particle swarm increases during the optimization process, and it is easy to fall into local optimization. In order to jump out of the local optimization and improve the diversity of the particle swarm, the optimization idea of ​​the Tianniu algorithm is used for improvement; Without knowing the exact location of the food, the longhorn beetle senses it through the antennae on both sides of its head. Since the antennae are at different distances from the food, the concentration of the odor sensed is also different. The longhorn beetle can move towards the food based on the concentration difference of the sensed odor and reach the food after multiple iterations. The mathematical expression is: In the formula, is a random unit vector, indicating the direction of the beetle's whiskers at any time; rands represents a random vector of (-1,1); Dim represents the spatial dimension; X t+1 represents the centroid position of the beetle after n iterations at time t+1; X l , X r X is the position of the left and right antennae of the longicorn beetle; t represents the centroid position of the beetle after n iterations at time t; f function is the objective function; δt is the iteration step length; sign is the sign function; In order to improve the global optimization of the particle swarm, the particle swarm algorithm formula is improved: v n i =wv n-1 i +c1r1(p n i -x n i )+c2r2(g n -x n i )+c3r3(p n i -g n ); In the formula, c3 is the variation factor; r3 is a random number between 0 and 1; The optimal positions of the current particle and the group are introduced in the formula, the positions of the two whiskers are regarded as the local optimum of the individual and the local optimum of the group, and the distance between the two whiskers represents the distance between the optimal positions of the individual and the group.

9. The Buck-Boost converter circuit for photovoltaic maximum power point tracking according to claim 8, characterized in that: The optimization convergence phase process of the optimized particle swarm algorithm BSPSO includes: After iteration, when the best position found by an individual particle swarm is better than the best position obtained by the group in the previous iteration, the particle swarm in the vicinity of the individual enters the optimization convergence state, switches the iterative algorithm, uses the simplified particle swarm PSO algorithm, removes the speed term, and speeds up the convergence speed; If we only consider the motion of the i-th particle, we can update the formula for the particle swarm by It can be changed to: x(n+1)=x(n)+v(n); After iterating the above formula, we get: The formula is a classical second-order differential equation without velocity terms; At this time, it can be transformed into a first-order differential equation: The particle swarm optimization equation without velocity term is simplified to: x n i =wx n-1 i +c1r1(p n i -x n i )+c2r2(g n -x n i ); After removing the velocity term, the particle swarm equation is reduced from a second-order differential equation to a first-order differential equation, which simplifies the iterative process and facilitates the analysis and control of the particle swarm evolution process.

10. The Buck-Boost converter circuit for photovoltaic maximum power point tracking according to claim 1, characterized in that: The adaptive parameter process of optimizing the particle swarm algorithm BSPSO includes: Adopt nonlinear strategy to update inertia weight w: Where w1 is the maximum value of the inertia weight w; w2 is the minimum value of the inertia weight w; h is the nonlinear coefficient; k m is the maximum number of iterations set; k is the current number of iterations; The individual learning factor c1 and the group learning factor c2 are updated using the trigonometric formula: The mutation factor c3 is updated using a nonlinear strategy: In the formula, c α =0.9, c β =0.45.

Images