MPPT-based photovoltaic hydrogen production power supply topology and control method

Abstract

The application discloses a photovoltaic hydrogen production power supply topology structure and a control method based on MPPT, relates to the technical field of renewable energy electrolysis hydrogen production, and mainly comprises a photovoltaic power module, a direct-current bus, a water electrolysis hydrogen production unit, an energy storage unit and a control unit. The photovoltaic power module is connected in parallel to the direct-current bus through independent DC / DC converters and cooperates with the automatic identification and scheduling function of the control unit to automatically adjust the system operation strategy when a module is added or fails to exit, can support plug and play, fault isolation and power reconstruction of the photovoltaic power module, limits the duty cycle change step of the DC / DC converter when the current rapidly fluctuates due to the change of photovoltaic output power, effectively controls the current change rate of the electrolytic cell stack, can protect the service life of the electrolytic cell, has high modularization access capacity, supports multi-source collaborative scheduling, and can realize bus voltage stabilization and system current sharing optimization.

Description

Technical Field

[0001] This invention relates to the field of renewable energy electrolysis hydrogen production technology, and in particular to a photovoltaic hydrogen production power supply topology and control method based on MPPT. Background Technology

[0002] Photovoltaic power generation, by directly converting solar energy into electricity, provides a renewable energy source for clean hydrogen production. With the transformation of the global energy structure and the advancement of carbon neutrality goals, photovoltaic-driven green hydrogen production technology is gradually becoming an important pathway for the deep utilization of renewable energy. However, existing photovoltaic hydrogen production systems have revealed several key technological bottlenecks during long-term operation.

[0003] First, traditional maximum power point tracking (MPPT) control algorithms generally aim to maximize photovoltaic power output, lacking dynamic adaptation considerations for the operating characteristics of downstream electrolyzers. In actual operation, frequent and drastic current fluctuations can accelerate the aging of electrolyzer electrode materials, significantly shortening equipment lifespan and compromising long-term operational reliability. Second, existing photovoltaic hydrogen production power structures mostly adopt centralized topologies, lacking flexible modular expansion capabilities. In the event of system expansion or partial failure, plug-and-play, scheduling isolation, and load balancing of power modules cannot be achieved, severely restricting system maintainability and scalability. Third, in application scenarios where multiple renewable DC power sources (such as photovoltaic, wind, and energy storage) are connected in parallel, traditional energy management solutions generally adopt static strategies, failing to fully coordinate the dynamic response capabilities and operating costs of each power source, resulting in low power distribution efficiency, large bus voltage fluctuations, and uneven load on power equipment.

[0004] Furthermore, in distributed photovoltaic module grid-connected hydrogen production systems, the bus cable layout is complex, and voltage unevenness and circulating current interference are prone to occur. If there is a lack of an effective current sharing regulation mechanism, the system stability risk will be further aggravated.

[0005] Therefore, how to construct a photovoltaic hydrogen production power supply topology and control method that can both protect the lifespan of the electrolyzer, have highly modular access capabilities, support multi-source collaborative scheduling, and achieve bus voltage stability and system current sharing optimization has become a key technical problem that urgently needs to be solved in this field. Summary of the Invention

[0006] The purpose of this invention is to solve the problems of large power fluctuations, limited electrolyzer lifespan, difficulty in module expansion, and weak multi-source power supply coordination in existing photovoltaic hydrogen production systems. It proposes a photovoltaic hydrogen production power supply topology and control method based on MPPT, which can take into account the protection of electrolyzer lifespan, has a high degree of modular access capability, supports multi-source collaborative scheduling, and can achieve bus voltage stability and system current sharing optimization.

[0007] The present invention achieves the above objectives through the following technical solutions:

[0008] A photovoltaic hydrogen production power supply topology based on MPPT includes several photovoltaic module strings, a DC / DC converter, a DC bus, a water electrolysis hydrogen production unit, an energy storage unit, and a control unit;

[0009] The photovoltaic module string and the corresponding independent DC / DC converter constitute a photovoltaic power module, and multiple photovoltaic power modules are connected in parallel to the DC bus.

[0010] The water electrolysis hydrogen production unit includes an electrolyzer stack, with both ends of the electrolyzer stack connected to the DC bus for receiving DC power to produce hydrogen through water electrolysis.

[0011] The energy storage unit is connected in parallel to the DC bus via a bidirectional DC / DC converter for energy buffering and power regulation.

[0012] The control unit includes:

[0013] The MPPT control subunit is used to control multiple DC / DC converters respectively, so that the DC / DC converters operate at their respective maximum power points according to the operating status of the corresponding photovoltaic module strings.

[0014] An electrolytic cell protection control subunit is used to limit the duty cycle change step of the DC / DC converter when the rate of change of the input current of the electrolytic cell stack exceeds the current change rate threshold caused by the change of photovoltaic output power, so as to control the rate of change of the current of the electrolytic cell stack to not exceed the current change rate threshold.

[0015] The energy management subunit is used to coordinate the charging and discharging state of the energy storage unit and / or schedule the output of other DC power sources based on the power difference between the photovoltaic output power and the power demand of the electrolytic cell stack when the energy storage unit and / or other DC power sources are connected in parallel on the DC bus, so as to achieve dynamic balance of the bus voltage.

[0016] The control unit is further configured to automatically identify the access status of the photovoltaic power modules. If a new photovoltaic power module is added, it will be automatically included in the control scheduling. When any photovoltaic power module fails and exits operation, the remaining photovoltaic power modules will automatically adjust their output to share the load and maintain stable system operation.

[0017] A further improvement of the present invention is that the DC bus adopts a voltage-current dual-sensing segmented current sharing coordination structure, including segmented bus cables, with voltage sampling nodes and current sampling nodes respectively set at both ends of each segment of bus cable;

[0018] The control unit includes multiple distributed current sharing regulators. Each current sharing regulator collects the voltage and current values ​​of the corresponding bus segment and controls the operating state of the DC / DC converter connected to that bus segment.

[0019] The control unit adjusts the output parameters of the corresponding current sharing controller through a closed-loop control algorithm based on the voltage difference and current imbalance between any two adjacent bus segments, thereby achieving balanced regulation of the output current of each photovoltaic power module or energy storage unit.

[0020] Each current sharing controller has a communication function to exchange parameters and update the regulation strategy with the control unit when the photovoltaic power module or energy storage unit is connected or disconnected.

[0021] A further improvement of the present invention is that: the electrolyzer protection and control subunit estimates the operating health factor of the electrolyzer stack based on the voltage, current, internal resistance, temperature and hydrogen yield of the electrolyzer stack using an artificial neural network algorithm (ANN), and dynamically adjusts the current change rate threshold according to the operating health factor.

[0022] A further improvement of the present invention is that: in each control cycle, the electrolytic cell protection control subunit constructs a loss function based on the difference between the current and previous cycle electrolytic cell stack input current, and uses an online gradient descent algorithm to iteratively update the duty cycle change step size of the DC / DC converter.

[0023] When the rate of change of the input current of the electrolytic cell stack exceeds the dynamically adjusted current change rate threshold, the duty cycle change step size of the DC / DC converter is automatically reduced.

[0024] A further improvement of the present invention is that the energy management subunit is configured with a resource scheduling evaluation model based on the unit power output cost, response time parameters and current available power of each DC power source;

[0025] Based on the resource scheduling evaluation model, an objective function for optimized scheduling is constructed. With the power demand of the electrolytic cell stack as a constraint, the output power of various DC power sources, including photovoltaic module strings, wind power generation, fuel cells, and supercapacitors, is dynamically allocated according to the objective function in each control cycle. Output power quota constraints are set during the allocation process to avoid any DC power source from operating under overload for a long time.

[0026] A further improvement of the present invention is that the energy management subunit adopts a model predictive control strategy, predicts the output power disturbance trend of the photovoltaic module string and the power demand of the electrolytic cell stack within a certain number of time steps in each control cycle, and solves the optimization control quantity based on the prediction results to adjust the charging and discharging power of the energy storage unit so that the DC bus voltage approaches the rated operating voltage of the electrolytic cell stack.

[0027] The objective function of the model predictive control strategy is as follows:

[0028] ;

[0029] In the formula, , The first Time, Number The energy storage charging and discharging power at any given moment; To predict the time domain length, dynamic adjustments are made based on power fluctuations; , To adjust the weighting factor; For the first The bus voltage predicted at any given time; The set operating voltage of the electrolytic cell stack; For the first The photovoltaic power variation predicted at any time.

[0030] A further improvement of the present invention is that each photovoltaic power module integrates a unique module identification information and a status broadcasting mechanism, and calculates the power capability parameters of the current maximum power point in real time, and periodically broadcasts module status data frames containing module identification information, operating status and power capability parameters to the control unit through the communication bus;

[0031] The control unit automatically identifies newly added photovoltaic power modules and dynamically maintains the module status mapping table by listening to module status data frames on the communication bus.

[0032] The control unit further employs a hash mapping algorithm or a distributed scheduling optimization mechanism to allocate corresponding MPPT scheduling resources based on the operating status and power capability parameters in the module status data frame, and constructs an MPPT control task queue according to adaptive scheduling priority to achieve real-time optimization of multi-module power output.

[0033] The control unit also includes a redundancy identification and fault tolerance mechanism. If no module status data frame of any photovoltaic power module is received continuously within a preset time window, it is marked as temporarily offline and the historical scheduling parameters of the photovoltaic power module are retained. When the photovoltaic power module resends the module status data frame, the control unit automatically identifies the reconnection status and restores the corresponding original scheduling resource allocation.

[0034] A control method for a photovoltaic hydrogen production power supply topology based on MPPT, the method comprising:

[0035] The system collects the output voltage, current, and operating status of the photovoltaic module string, and controls each DC / DC converter through the MPPT control subunit to make each photovoltaic power module work at its own maximum power point.

[0036] The energy management subunit uses model predictive control strategies to coordinate the charging and discharging states of the energy storage unit, or constructs a resource scheduling evaluation model based on the difference between the photovoltaic output power and the power demand of the electrolyzer stack, and schedules the output power of other DC power sources connected in parallel to the DC bus to achieve dynamic balance of the bus power.

[0037] The voltage, current, internal resistance, temperature and hydrogen yield of the electrolyzer stack are collected. The operating health factor of the electrolyzer stack is estimated using the artificial neural network algorithm (ANN). The current change rate threshold is dynamically set based on the operating health factor.

[0038] In each control cycle, the electrolytic cell protection control subunit constructs a loss function based on the difference in input current of the electrolytic cell stack between adjacent cycles, iteratively updates the duty cycle change step size of the DC / DC converter using an online gradient descent algorithm, and automatically shrinks the duty cycle change step size when the rate of change of input current of the electrolytic cell stack exceeds the set current change rate threshold.

[0039] The photovoltaic power module calculates the power capability parameters of the current maximum power point in real time and periodically broadcasts module status data frames through the communication bus, including module identification information, operating status and power capability parameters. The control unit receives the module status data frames, automatically identifies the access or exit status of the photovoltaic power module, updates the scheduling resource allocation, and restores the scheduling status of the faulty module through a redundancy identification mechanism.

[0040] A further improvement of the present invention is that: the energy management subunit constructs a resource scheduling evaluation model based on the unit power output cost, response time parameters and current available power of each DC power source, and dynamically allocates the output power of various DC power sources, including photovoltaic module strings, wind power generation, fuel cells, and supercapacitors;

[0041] The objective function of the resource scheduling evaluation model is expressed as:

[0042] ;

[0043] In the formula, Indicates assignment to the first The output power of a DC power supply; The total number of DC power supplies participating in the dispatch; This is an adjustment coefficient related to the output cost weight. For the first The output power of each DC power supply is Cost function per unit power output at time; The adjustment coefficient is related to the response time weight. For the first The dynamic response time of a DC power supply, in seconds; This represents the total load power currently required by the electrolytic cell stack.

[0044] A further improvement of the present invention is that the electrolyzer protection and control subunit inputs the voltage, current, internal resistance, temperature, and hydrogen yield of the electrolyzer stack into a pre-trained artificial neural network model. The artificial neural network model is a three-layer feedforward structure, with the input layer containing the above five parameters and the output layer being a normalized operating health factor. It is used to characterize the current operating aging state of the electrolytic cell stack;

[0045] Based on the aforementioned operational health factors Dynamically calculate the maximum allowable rate of change threshold of current. The formula is:

[0046] ;

[0047] In the formula, The maximum rate of change of current initially set for the system; , These are adjustable weighting coefficients; , They are respectively The operating health factor and current change rate threshold at any given time;

[0048] Within each control cycle, the input current difference of the electrolytic cell stack is calculated. And construct the loss function. ,in, , , , Input current to the electrolytic cell stack for adjacent cycles;

[0049] The duty cycle step size of the DC / DC converter is iteratively updated using an online gradient descent algorithm. The update rules are as follows:

[0050] ;

[0051] when At that time, the step size of the forced contraction duty cycle change is:

[0052] ;

[0053] In the formula, The learning rate; The contraction factor is dynamically adjusted based on the extent of the over-limit.

[0054] The MPPT control subunit employs a maximum power point tracking strategy that includes feedback on operational health factors. The objective function is as follows:

[0055] ;

[0056] In the formula, The optimized maximum power point target voltage; Indicates voltage Find the voltage point within the space that maximizes the expression within the parentheses; Let be the efficiency function of the photovoltaic module; Lifetime protection weighting factor;

[0057] The MPPT control subunit outputs the target voltage according to the objective function. The system adjusts the operating voltage of the photovoltaic module strings in real time to balance the photovoltaic output efficiency with the protection requirements of the electrolyzer stack lifespan, and calculates the power capacity parameters at the current maximum power point in real time. The formula is: ,in This represents the output current at the current voltage.

[0058] The beneficial effects of this invention are as follows: Each photovoltaic power module is connected in parallel to the DC bus through an independent DC / DC converter, and with the automatic identification and scheduling function of the control unit, it can support plug-and-play, fault isolation, and power reconfiguration of the photovoltaic power modules, providing good flexibility and ease of maintenance; when a module is added or fails, the system operation strategy is automatically adjusted to maintain overall power supply continuity and improve system availability and redundancy fault tolerance; when changes in photovoltaic output power cause rapid current fluctuations, the duty cycle change step of the DC / DC converter is limited, effectively controlling the current change rate of the electrolytic cell stack and preventing overheating of the electrolytic cell due to frequent current surges. This system addresses phenomena such as increased polarization, thereby extending its service life and improving system operational stability. Through the MPPT control subunit, it controls the DC / DC converters corresponding to multiple photovoltaic module strings, ensuring each photovoltaic power module always operates at its maximum power point, improving photovoltaic energy conversion efficiency and enhancing the system's power generation capacity under different lighting conditions. When the energy storage unit and / or other DC power sources are connected to the bus, it can dynamically adjust the energy storage charging and discharging state or schedule other power outputs based on the real-time difference between photovoltaic output and electrolytic cell power demand, achieving stable control of the DC bus voltage and enhancing the system's adaptability to load fluctuations and the uncertainties of renewable energy. Attached Figure Description

[0059] To more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings used in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort. Wherein:

[0060] Figure 1 This is a schematic diagram of the topology in an embodiment of the present invention;

[0061] Figure 2 This is a flowchart of the method in an embodiment of the present invention;

[0062] Figure 3 This is a flowchart of the combined method of MPPT control and electrolytic cell protection coupling control in an embodiment of the present invention. Detailed Implementation

[0063] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of the present invention. All other embodiments obtained by those skilled in the art based on the described embodiments of the present invention are within the scope of protection of the present invention.

[0064] like Figure 1 As shown, this is an embodiment of the present invention, which provides a photovoltaic hydrogen production power supply topology based on MPPT, including several photovoltaic module strings, a DC / DC converter, a DC bus, a water electrolysis hydrogen production unit, an energy storage unit, a control unit, etc.

[0065] Photovoltaic power modules: Photovoltaic module (PV) consists of strings of PV modules and corresponding independent DC / DC converters. Each string of PV modules is connected to the DC bus via its own DC / DC converter. Each PV power module is equipped with independent MPPT control, which adjusts the DC / DC converter duty cycle in real time to track the module's maximum power point. The use of multiple low-power DC / DC converters instead of a single high-power converter enables a modular design on the PV side. This modular topology allows each PV string to output and be protected independently: when the power of one PV string decreases due to shading or a fault, the remaining PV strings are unaffected and can still operate at their respective maximum power points. Furthermore, the system capacity can be flexibly expanded by adding or removing PV power modules, providing plug-and-play functionality.

[0066] The water electrolysis hydrogen production unit includes an electrolyzer stack, which is composed of several individual water electrolysis cells connected in series. The positive and negative terminals of the electrolyzer stack are connected to the output terminals of each DC / DC converter via a DC bus, receiving power from the DC bus to perform the water electrolysis reaction to produce hydrogen. To protect the electrolyzer, the topology of this invention is designed to ensure that the voltage across the electrolyzer stack is always maintained within a safe operating range, and the ripple of the bus current / voltage is reduced through control strategies. If necessary, an RC filter can be connected in parallel across the electrolyzer or an isolated DC / DC converter can be used to further improve the output quality. However, in the topology of this embodiment, due to multi-source coordinated control, the electrolyzer can already obtain a DC power supply with extremely low ripple.

[0067] Energy storage unit (or other DC power source): Preferably includes an energy storage battery and its bidirectional DC / DC converter. The energy storage unit is connected to the DC bus and is used to store energy when there is surplus photovoltaic power and release energy when there is insufficient photovoltaic power, thereby smoothing out photovoltaic power fluctuations and ensuring continuous and stable power supply to the electrolyzer. The bidirectional DC / DC converter controls the charging and discharging current of the battery to support and regulate the DC bus voltage; on the other hand, it can coordinate with the photovoltaic-side MPPT through the upper-level control system to provide instantaneous power compensation or discharge to the electrolyzer during dynamic processes. In addition to the battery, the DC bus of this invention can also be connected in parallel to other DC power sources, such as wind power rectifier output, supercapacitors, DC distribution networks, etc., to form a DC microgrid architecture. Multiple energy sources are connected to the common bus through their respective power conversion interfaces and are uniformly dispatched by the energy management system to supply power to the electrolyzer. When photovoltaic power is temporarily excessive, the surplus electricity can be stored in the battery or supplied to other loads; conversely, when there is insufficient sunlight, energy storage or other power sources make up for the power gap, thereby significantly improving the power supply stability and availability of the hydrogen production system.

[0068] DC Bus: Composed of conductor buses, connecting the DC / DC outputs of the above units in parallel. The bus voltage is determined by the instantaneous operating voltage of the electrolytic cell stack (or by a reference set by the battery converter, etc.), and in embodiments of this invention, it is typically maintained at a level close to the rated voltage of the electrolytic cell. Considering long-distance transmission and the need to reduce losses, the bus voltage can be designed to be higher (e.g., several hundred volts DC); each DC / DC converter adjusts its output according to the bus voltage to inject current into or draw current from the bus. The bus structure makes the entire system a star-parallel topology, with the photovoltaic side, energy storage side, and load (electrolytic cell) each connected to the bus via electronic converters. Electrical isolation and decoupling between modules are achieved by the control of their respective converters. If necessary, a DC / DC converter with a high-frequency isolation transformer can be selected to ensure electrical isolation between the photovoltaic and the electrolytic cell, improving safety and immunity.

[0069] Control unit: Includes MPPT control subunit, electrolyzer protection control subunit, and energy management subunit. The MPPT control subunit controls multiple DC / DC converters, ensuring they operate at their respective maximum power points based on the operating status of the corresponding photovoltaic module strings. The electrolyzer protection control subunit limits the duty cycle change step of the DC / DC converters when changes in photovoltaic output power cause the input current of the electrolyzer stack to exceed the current change rate threshold, thus controlling the current change rate of the electrolyzer stack to not exceed the current change rate threshold. When energy storage units and / or other DC power supplies are connected in parallel to the DC bus, the energy management subunit coordinates the charging and discharging states of the energy storage units and / or schedules the output of other DC power supplies based on the power difference between the photovoltaic output power and the power demand of the electrolyzer stack, achieving dynamic balance of the bus voltage.

[0070] The control unit is further configured to automatically identify the access status of photovoltaic power modules. If a new photovoltaic power module is added, it will be automatically included in the control scheduling. When any photovoltaic power module fails and exits operation, the remaining photovoltaic power modules will automatically adjust their output to share the load and maintain stable system operation.

[0071] The above topology fully embodies the design philosophy of modularity and multi-source collaboration. By operating multiple power conversion modules in parallel, it achieves an "N+1" redundancy configuration. Even if one module fails, the others can still operate normally, significantly improving system reliability and maintainability. Simultaneously, while ensuring electrical isolation and independent control of each module, it achieves plug-and-play photovoltaic power generation units. New modules are automatically identified and put into operation by the control unit after being connected to the bus, without requiring major modifications to the existing system. This architecture facilitates future expansion of photovoltaic capacity or the integration of new DC power sources (such as wind power).

[0072] In one embodiment, the DC bus adopts a voltage-current dual-sensing segmented current sharing coordination structure to solve the problem of uneven bus current distribution and increased circulating current risk caused by physical wiring asymmetry and load concentration when photovoltaic power modules and energy storage units are connected in parallel.

[0073] The DC bus is divided into several independent bus cables. Each bus cable has a voltage sampling node and a current sampling node at both ends. Each bus cable has two output ports corresponding to two photovoltaic power modules or energy storage units at both ends.

[0074] Each busbar segment is equipped with a distributed current sharing controller, whose functions include:

[0075] Sampling: Used to collect the voltage values ​​at both ends of this section of bus cable in real time. , and the current value through the cable ;

[0076] Control: Embedded microcontroller with local closed-loop adjustment algorithm;

[0077] Execution: Controls the operating status of the DC / DC converter between this bus section and the connected photovoltaic power module, and adjusts its output power;

[0078] Communication: Maintains periodic communication with the control unit to synchronize module status, power quotas, and current sharing commands.

[0079] The control unit determines the voltage difference between any two adjacent busbar segments based on the voltage difference between them. ) and the degree of current imbalance ( The target adjustment is calculated, and the output parameters of the corresponding flow-sharing controller are adjusted using a closed-loop control algorithm (such as proportional-integral PI control). The target of the control algorithm is: ;in, Number of busbar segments , The weighting factor is used to adjust the relative priority of voltage consistency and current balance.

[0080] When a photovoltaic power module or energy storage unit is added or removed, the corresponding module broadcasts the status change information on the communication bus. Each current sharing controller obtains this information through the communication interface and reports the access status to the control unit. The control unit automatically updates the regulation strategy based on the current system topology, including:

[0081] Reset the target voltage difference range for each section of the busbar;

[0082] Assign new flow equalization control priorities;

[0083] Boundary conditions are issued to each flow equalization controller.

[0084] Each current sharing controller can independently execute voltage closed-loop and current sharing control tasks locally based on the issued target and the voltage and current sampling results of the current segment. This enables distributed current sharing regulation and adaptive coordination between bus segments without the need for centralized real-time control, effectively preventing voltage drops, current overshoots, or circulating currents caused by local load changes.

[0085] This implementation method is suitable for large-scale parallel power supply scenarios containing multiple photovoltaic power modules and energy storage modules, and has good scalability and dynamic stability.

[0086] In one embodiment, the electrolyzer protection and control subunit estimates the operational health factors of the electrolyzer stack based on the voltage, current, internal resistance, temperature, and hydrogen yield of the electrolyzer stack using an artificial neural network algorithm (ANN). And dynamically adjust the current change rate threshold according to the operating health factors. The formula is:

[0087] ;

[0088] In the formula, The maximum rate of change of current initially set for the system; , These are adjustable weighting coefficients; , They are respectively The operating health factor and current change rate threshold at any given time;

[0089] In each control cycle, the electrolytic cell protection control subunit determines the current based on the difference between the current and previous cycle's electrolytic cell stack input current. Constructing the loss function The duty cycle change step size of the DC / DC converter is iteratively updated using an online gradient descent algorithm. When the rate of change of the input current to the electrolytic cell exceeds the dynamically adjusted current rate of change threshold, the duty cycle change step size of the DC / DC converter is automatically reduced. The MPPT control subunit calculates the optimized maximum power point target voltage based on feedback from the operating health factor. According to the target voltage The operating voltage of the photovoltaic module string is adjusted in real time to balance the photovoltaic output efficiency and the life protection requirements of the electrolytic cell stack.

[0090] In one embodiment, the energy management subunit is configured with a resource scheduling evaluation model based on the unit power output cost, response time parameters, and current available power of each DC power source. Based on the resource scheduling evaluation model, an objective function for optimizing scheduling is constructed, with the power demand of the electrolyzer stack as a constraint. In each control cycle, the output power of various DC power sources, including photovoltaic module strings, wind power generation, fuel cells, and supercapacitors, is dynamically allocated according to the objective function. Output power quota constraints are set during the allocation process to avoid any DC power source from operating under overload for a long time.

[0091] Optionally, in another embodiment, the energy management subunit may also employ a model predictive control strategy to predict the output power disturbance trend of the photovoltaic module string and the power demand of the electrolytic cell stack within a certain number of time steps in each control cycle, and solve the optimization control quantity based on the prediction results to adjust the charging and discharging power of the energy storage unit so that the DC bus voltage approaches the rated operating voltage of the electrolytic cell stack.

[0092] The objective function of the model predictive control strategy is as follows:

[0093] ;

[0094] In the formula, , The first Time, Number The energy storage charging and discharging power at any given moment; To predict the time domain length, dynamic adjustments are made based on power fluctuations; , To adjust the weighting factor; For the first The bus voltage predicted at any given time; The set operating voltage of the electrolytic cell stack; For the first The photovoltaic power variation predicted at any time.

[0095] Specifically, as one implementation of the present invention, each photovoltaic power module integrates unique module identification information and a status broadcasting mechanism, and calculates the power capability parameters of the current maximum power point in real time. It periodically broadcasts module status data frames containing module identification information, operating status and power capability parameters to the control unit through the communication bus.

[0096] The control unit automatically identifies newly added photovoltaic power modules and dynamically maintains the module status mapping table by listening to module status data frames on the communication bus;

[0097] The control unit further employs a hash mapping algorithm or a distributed scheduling optimization mechanism to allocate corresponding MPPT scheduling resources based on the operating status and power capability parameters in the module status data frame, and constructs an MPPT control task queue according to adaptive scheduling priority to achieve real-time optimization of multi-module power output;

[0098] The control unit also includes a redundancy identification and fault tolerance mechanism. If no module status data frame is continuously received from any photovoltaic power module within a preset time window, it is marked as temporarily offline and the historical scheduling parameters of the photovoltaic power module are retained. When the photovoltaic power module resends the module status data frame, the control unit automatically identifies the reconnection status and restores the corresponding original scheduling resource allocation.

[0099] like Figure 2 , Figure 3 As shown, this is an embodiment of the present invention, which provides a control method for a photovoltaic hydrogen production power supply topology based on MPPT, including:

[0100] The system collects the output voltage, current, and operating status of the photovoltaic module string, and controls each DC / DC converter through the MPPT control subunit to make each photovoltaic power module work at its own maximum power point.

[0101] The energy management subunit employs model predictive control strategies to coordinate the charging and discharging states of the energy storage unit, or constructs a resource scheduling evaluation model based on the difference between the photovoltaic output power and the electrolyzer stack power demand, scheduling the output power of other DC power sources connected in parallel to the DC bus to achieve dynamic balance of the bus power. Specifically, the energy management subunit constructs a resource scheduling evaluation model based on the unit power output cost, response time parameters, and current available power of each DC power source, dynamically allocating the output power of various DC power sources, including photovoltaic module strings, wind power generation, fuel cells, and supercapacitors. The objective function of the resource scheduling evaluation model is expressed as:

[0102] ;

[0103] In the formula, Indicates assignment to the first The output power of a DC power supply; The total number of DC power supplies participating in the dispatch; This is an adjustment coefficient related to the output cost weight. For the first The output power of each DC power supply is Cost function per unit power output at time; The adjustment coefficient is related to the response time weight. For the first The dynamic response time of a DC power supply, in seconds; This represents the total load power currently required by the electrolytic cell stack.

[0104] The voltage, current, internal resistance, temperature, and hydrogen yield of the electrolyzer stack are collected. An artificial neural network (ANN) algorithm is used to estimate the operational health factors of the electrolyzer stack. Specifically, the electrolyzer protection and control subunit inputs the voltage, current, internal resistance, temperature, and hydrogen yield of the electrolyzer stack into a pre-trained ANN model. The ANN model has a three-layer feedforward structure; the input layer contains the above five parameters, and the output layer is the normalized operational health factor. It is used to characterize the current operating aging state of the electrolytic cell stack;

[0105] Based on operational health factors Dynamically adjust the maximum allowable rate of change threshold of current. The formula is:

[0106] ;

[0107] In the formula, The maximum rate of change of current initially set for the system; , These are adjustable weighting coefficients; , They are respectively The operating health factor and current change rate threshold at any given time;

[0108] Within each control cycle, the input current difference of the electrolytic cell stack is calculated. And construct the loss function. ,in, , , , Input current to the electrolytic cell stack for adjacent cycles;

[0109] The duty cycle step size of the DC / DC converter is iteratively updated using an online gradient descent algorithm. The update rules are as follows:

[0110] ;

[0111] when At that time, the step size of the forced contraction duty cycle change is:

[0112] ;

[0113] In the formula, The learning rate; The contraction factor is dynamically adjusted based on the extent of the over-limit.

[0114] The photovoltaic power module calculates the power capability parameters of the current maximum power point in real time. Specifically, it adopts a maximum power point tracking strategy that includes feedback on operational health factors. The objective function is as follows:

[0115] ;

[0116] In the formula, The optimized maximum power point target voltage; Indicates voltage Find the voltage point within the space that maximizes the expression within the parentheses; Let be the efficiency function of the photovoltaic module; Lifetime protection weighting factor;

[0117] The MPPT control subunit outputs the target voltage based on the objective function. The system adjusts the operating voltage of the photovoltaic module strings in real time to balance the photovoltaic output efficiency with the protection requirements of the electrolyzer stack lifespan, and calculates the power capacity parameters at the current maximum power point in real time. The formula is: ,in This represents the output current at the current voltage.

[0118] Each photovoltaic power module is written with module identification information (unique ID, such as MAC, SN, or UUID) during manufacturing or initialization, and is configured with an embedded MCU (such as STM32 or ESP32) that periodically broadcasts module status data frames, including module identification information, operating status (running, standby, fault), and power capability parameters (excluding...). External parameters may also include voltage range, current limiting value, connection timestamp, etc. The control unit receives module status data frames, automatically identifies the access or exit status of photovoltaic power modules, updates the allocation of scheduling resources, and restores the scheduling status of faulty modules through a redundancy identification mechanism.

[0119] In summary, this invention addresses both system structure and control strategy. By constructing a modular, scalable, and multi-source collaborative photovoltaic hydrogen production power topology, and integrating control mechanisms such as electrolyzer lifetime protection, dynamic MPPT scheduling, energy storage coordinated control, and bus current sharing management, the invention achieves efficient utilization of photovoltaic energy, optimized lifespan of the water electrolysis system, flexible scheduling of power modules, dynamic balance of multi-source power supply, and safe and reliable system operation. This significantly improves the overall performance, stability, and engineering applicability of the photovoltaic hydrogen production system.

[0120] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any person skilled in the art can easily conceive of various variations or substitutions within the technical scope disclosed in this application, and these should all be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

Claims

1. A photovoltaic hydrogen production power supply topology based on MPPT, comprising several photovoltaic module strings, a DC / DC converter, a DC bus, a water electrolysis hydrogen production unit, an energy storage unit, and a control unit, characterized in that: The photovoltaic module string and the corresponding independent DC / DC converter constitute a photovoltaic power module, and multiple photovoltaic power modules are connected in parallel to the DC bus. The water electrolysis hydrogen production unit includes an electrolyzer stack, with both ends of the electrolyzer stack connected to the DC bus for receiving DC power to produce hydrogen through water electrolysis. The energy storage unit is connected in parallel to the DC bus via a bidirectional DC / DC converter for energy buffering and power regulation. The control unit includes: The MPPT control subunit is used to control multiple DC / DC converters respectively, so that the DC / DC converters operate at their respective maximum power points according to the operating status of the corresponding photovoltaic module strings. An electrolytic cell protection control subunit is used to limit the duty cycle change step of the DC / DC converter when the rate of change of the input current of the electrolytic cell stack exceeds the current change rate threshold caused by the change of photovoltaic output power, so as to control the rate of change of the current of the electrolytic cell stack to not exceed the current change rate threshold. The energy management subunit is used to coordinate the charging and discharging state of the energy storage unit and / or schedule the output of other DC power sources based on the power difference between the photovoltaic output power and the power demand of the electrolytic cell stack when the energy storage unit and / or other DC power sources are connected in parallel on the DC bus, so as to achieve dynamic balance of the bus voltage. The control unit is further configured to automatically identify the access status of the photovoltaic power modules. If a new photovoltaic power module is added, it will be automatically included in the control scheduling. When any photovoltaic power module fails and exits operation, the remaining photovoltaic power modules will automatically adjust their output to share the load and maintain stable system operation.

2. The photovoltaic hydrogen production power supply topology based on MPPT according to claim 1, characterized in that: The DC bus adopts a voltage-current dual-sensor segmented current sharing coordination structure, including segmented bus cables, with voltage sampling nodes and current sampling nodes respectively set at both ends of each segment of bus cable. The control unit includes multiple distributed current sharing regulators. Each current sharing regulator collects the voltage and current values ​​of the corresponding bus segment and controls the operating state of the DC / DC converter connected to that bus segment. The control unit adjusts the output parameters of the corresponding current sharing controller through a closed-loop control algorithm based on the voltage difference and current imbalance between any two adjacent bus segments, thereby achieving balanced regulation of the output current of each photovoltaic power module or energy storage unit. Each current sharing controller has a communication function to exchange parameters and update the regulation strategy with the control unit when the photovoltaic power module or energy storage unit is connected or disconnected.

3. The photovoltaic hydrogen production power supply topology based on MPPT according to claim 1, characterized in that: The electrolyzer protection and control subunit estimates the operating health factors of the electrolyzer stack based on the voltage, current, internal resistance, temperature and hydrogen yield of the electrolyzer stack using an artificial neural network algorithm (ANN), and dynamically adjusts the current change rate threshold according to the operating health factors.

4. The photovoltaic hydrogen production power supply topology based on MPPT according to claim 3, characterized in that: In each control cycle, the electrolytic cell protection control subunit constructs a loss function based on the difference between the current and previous cycle's electrolytic cell stack input current, and uses an online gradient descent algorithm to iteratively update the duty cycle change step size of the DC / DC converter. When the rate of change of the input current of the electrolytic cell stack exceeds the dynamically adjusted current change rate threshold, the duty cycle change step size of the DC / DC converter is automatically reduced.

5. The photovoltaic hydrogen production power supply topology based on MPPT according to claim 1, characterized in that: The energy management subunit is equipped with a resource scheduling evaluation model built based on the unit power output cost, response time parameters, and current available power of each DC power source. Based on the resource scheduling evaluation model, an objective function for optimized scheduling is constructed. With the power demand of the electrolytic cell stack as a constraint, the output power of various DC power sources, including photovoltaic module strings, wind power generation, fuel cells, and supercapacitors, is dynamically allocated according to the objective function in each control cycle. Output power quota constraints are set during the allocation process to avoid any DC power source from operating under overload for a long time.

6. The photovoltaic hydrogen production power supply topology based on MPPT according to claim 1, characterized in that: The energy management subunit adopts a model predictive control strategy to predict the output power disturbance trend of the photovoltaic module string and the power demand of the electrolytic cell stack within several time steps in each control cycle, and solves the optimization control quantity based on the prediction results to adjust the charging and discharging power of the energy storage unit so that the DC bus voltage approaches the rated operating voltage of the electrolytic cell stack. The objective function of the model predictive control strategy is as follows: ； In the formula, , The first Time, Number The energy storage charging and discharging power at any given moment; To predict the time domain length, dynamic adjustments are made based on power fluctuations; , To adjust the weighting factor; For the first The bus voltage predicted at any given time; The set operating voltage of the electrolytic cell stack; For the first The photovoltaic power variation is predicted at any time.

7. The photovoltaic hydrogen production power supply topology based on MPPT according to claim 1, characterized in that: Each photovoltaic power module integrates unique module identification information and a status broadcasting mechanism, and calculates the power capability parameters of the current maximum power point in real time. It periodically broadcasts module status data frames containing module identification information, operating status and power capability parameters to the control unit through the communication bus. The control unit automatically identifies newly added photovoltaic power modules and dynamically maintains the module status mapping table by listening to module status data frames on the communication bus. The control unit further employs a hash mapping algorithm or a distributed scheduling optimization mechanism to allocate corresponding MPPT scheduling resources based on the operating status and power capability parameters in the module status data frame, and constructs an MPPT control task queue according to adaptive scheduling priority to achieve real-time optimization of multi-module power output. The control unit also includes a redundancy identification and fault tolerance mechanism. If no module status data frame of any photovoltaic power module is received continuously within a preset time window, it is marked as temporarily offline and the historical scheduling parameters of the photovoltaic power module are retained. When the photovoltaic power module resends the module status data frame, the control unit automatically identifies the reconnection status and restores the corresponding original scheduling resource allocation.

8. A control method applied to the photovoltaic hydrogen production power supply topology based on MPPT as described in any one of claims 1-7, characterized in that: The method includes: The system collects the output voltage, current, and operating status of the photovoltaic module string, and controls each DC / DC converter through the MPPT control subunit to make each photovoltaic power module work at its own maximum power point. The energy management subunit uses model predictive control strategies to coordinate the charging and discharging states of the energy storage unit, or constructs a resource scheduling evaluation model based on the difference between the photovoltaic output power and the power demand of the electrolyzer stack, and schedules the output power of other DC power sources connected in parallel to the DC bus to achieve dynamic balance of the bus power. The voltage, current, internal resistance, temperature and hydrogen yield of the electrolyzer stack are collected. The operating health factor of the electrolyzer stack is estimated using the artificial neural network algorithm (ANN). The current change rate threshold is dynamically set based on the operating health factor. In each control cycle, the electrolytic cell protection control subunit constructs a loss function based on the difference in input current of the electrolytic cell stack between adjacent cycles, iteratively updates the duty cycle change step size of the DC / DC converter using an online gradient descent algorithm, and automatically shrinks the duty cycle change step size when the rate of change of input current of the electrolytic cell stack exceeds the set current change rate threshold. The photovoltaic power module calculates the power capability parameters of the current maximum power point in real time and periodically broadcasts module status data frames through the communication bus, including module identification information, operating status and power capability parameters. The control unit receives the module status data frames, automatically identifies the access or exit status of the photovoltaic power module, updates the scheduling resource allocation, and restores the scheduling status of the faulty module through a redundancy identification mechanism.

9. The control method according to claim 8, characterized in that: The energy management subunit constructs a resource scheduling evaluation model based on the unit power output cost, response time parameters, and current available power of each DC power source, and dynamically allocates the output power of various DC power sources, including photovoltaic module strings, wind power generation, fuel cells, and supercapacitors. The objective function of the resource scheduling evaluation model is expressed as: ； In the formula, Indicates assignment to the first The output power of a DC power supply; The total number of DC power supplies participating in the dispatch; This is an adjustment coefficient related to the output cost weight. For the first The output power of each DC power supply is Cost function per unit power output at time; The adjustment coefficient is related to the response time weight. For the first The dynamic response time of a DC power supply, in seconds; This represents the total load power currently required by the electrolytic cell stack.

10. The control method according to claim 8, characterized in that: The electrolyzer protection and control subunit inputs the voltage, current, internal resistance, temperature, and hydrogen yield of the electrolyzer stack into a pre-trained artificial neural network model. The artificial neural network model is a three-layer feedforward structure, with the input layer containing the above five parameters and the output layer being a normalized operating health factor. It is used to characterize the current operating aging state of the electrolytic cell stack; Based on the aforementioned operational health factors Dynamically calculate the maximum allowable rate of change threshold of current. The formula is: ； In the formula, The maximum rate of change of current initially set for the system; , These are adjustable weighting coefficients; , They are respectively The operating health factor and current change rate threshold at any given time; Within each control cycle, the input current difference of the electrolytic cell stack is calculated. And construct the loss function. ,in, , , , Input current to the electrolytic cell stack for adjacent cycles; The duty cycle step size of the DC / DC converter is iteratively updated using an online gradient descent algorithm. The update rules are as follows: ； when At that time, the step size of the forced contraction duty cycle change is: ； In the formula, The learning rate; The contraction factor is dynamically adjusted based on the extent of the over-limit. The MPPT control subunit employs a maximum power point tracking strategy that includes feedback on operational health factors. The objective function is as follows: ； In the formula, The optimized maximum power point target voltage; Indicates voltage Find the voltage point within the space that maximizes the expression within the parentheses; Let be the efficiency function of the photovoltaic module; Lifetime protection weighting factor; The MPPT control subunit outputs the target voltage according to the objective function. The system adjusts the operating voltage of the photovoltaic module strings in real time to balance the photovoltaic output efficiency with the protection requirements of the electrolyzer stack lifespan, and calculates the power capacity parameters at the current maximum power point in real time. The formula is: ,in This represents the output current at the current voltage.

Images