Distributed cooperative control system for photovoltaic hydrogen production direct current micro-grid

Abstract

The invention discloses a distributed cooperative control system for a photovoltaic hydrogen production direct-current micro-grid, and relates to the technical field of new energy power generation and power system control, and the system comprises a photovoltaic conversion module which is composed of a photovoltaic array and a direct-current boost converter and converts solar energy into electric energy; the hydrogen energy storage module is composed of an electrolytic cell and the like and achieves energy conversion and storage. The cooperative communication module receives and transmits the key state information and judges a working mode; the distributed optimization control module realizes distributed optimization and cooperative control, and solves and outputs an optimal switch control signal; an integrated cooperative control framework is constructed, a distributed control framework is used for replacing a traditional centralized control framework, the defects of the distributed control framework are avoided, and unification of all modules is achieved; meanwhile, an improved algorithm is introduced, a cost function is incorporated into multiple targets, the priority is scientifically coordinated, the strategy can be dynamically adjusted during fluctuation, waste is reduced, loss is reduced, and the multi-aspect performance of the microgrid is improved.

Description

Technical Field

[0001] This invention relates to the field of new energy power generation and power system control technology, specifically a distributed collaborative control system for photovoltaic hydrogen production DC microgrids. Background Technology

[0002] With the rapid development of renewable energy, photovoltaic (PV) DC microgrids, due to their high efficiency and flexibility, have been widely used in the field of new energy utilization and have become an important technological direction for promoting energy structure transformation. PV DC microgrids can directly connect to DC loads and DC-type energy storage devices, reducing losses in the power conversion process and improving energy utilization efficiency. However, the output characteristics of photovoltaic power generation are highly dependent on climate and environmental conditions such as solar irradiance and ambient temperature, exhibiting significant randomness and intermittency. This makes it difficult to achieve real-time matching between the output power on the generation side and the demand on the load side, easily leading to fluctuations in DC bus voltage. This not only affects the normal operation of the load but may also result in excess energy not being effectively utilized, leading to energy waste. To mitigate the impact of photovoltaic volatility on system stability, the industry is gradually exploring the integration of hydrogen energy storage systems with PV DC microgrids to construct a closed-loop energy system of "power generation—hydrogen production—energy storage." Excess electricity is converted into hydrogen energy for storage through an electrolyzer. When sunlight is insufficient, the system balance is maintained by relying on the external power grid or hydrogen-to-electricity conversion, thereby improving the operational stability and energy utilization efficiency of the PV DC microgrid.

[0003] Traditional photovoltaic (PV) DC microgrids often employ centralized energy management systems for power distribution and control. This architecture requires unified modeling and centralized solution of the global system state. As the system scales up or the number of nodes increases, it faces problems such as high communication latency, poor scalability, and the risk of single points of failure. If the central controller fails, the entire system may be paralyzed, making it difficult to meet the real-time control requirements of large-scale distributed energy systems. Furthermore, existing control methods often neglect the switching losses and dynamic constraints of power converters, focusing only on voltage and power balance, leading to increased equipment losses and decreased economic efficiency and reliability during system operation. In PV-hydrogen integrated systems, existing solutions generally employ hierarchical or centralized control, lacking comprehensive optimization of economics, equipment losses, and equipment lifespan. The dynamic characteristics of each subsystem differ significantly, and the collaborative control process is complex, making it difficult to achieve globally optimal operation. In addition, the computational complexity of traditional centralized economic model predictive control architecture increases exponentially with system scale, and communication load and latency issues become significant, making it unable to quickly respond to dynamic scenarios such as sudden changes in illumination and load fluctuations, further limiting the efficient operation of PV-hydrogen production DC microgrids. Summary of the Invention

[0004] The purpose of this invention is to overcome the shortcomings of existing technologies and provide a distributed collaborative control system for photovoltaic hydrogen production DC microgrids. This system converts solar energy into electrical energy through a photovoltaic conversion module, utilizes a hydrogen energy storage module for energy conversion and storage, and a collaborative communication module receives and transmits key system status information in real time to determine the system's operating mode. A distributed optimization control module realizes distributed economic optimization and collaborative control of the photovoltaic and hydrogen production modules. Through rolling optimization control, with the goal of achieving the best overall economic benefits of the system, an improved MINLP algorithm is used to solve for the optimal switching control signal.

[0005] To solve the above-mentioned technical problems, the present invention provides the following technical solution: a distributed collaborative control system for photovoltaic hydrogen production DC microgrids, the system comprising:

[0006] Photovoltaic conversion module: Composed of a photovoltaic array and a DC-DC boost converter, used to convert solar energy into electrical energy. The operating parameters of the photovoltaic array are affected by environmental conditions, and the DC-DC boost converter is regulated by a switch control signal.

[0007] Hydrogen energy storage module: It consists of an electrolyzer, a DC-DC step-down converter and a hydrogen storage tank, and is used for energy conversion and storage. The electrolyzer receives electrical energy to produce hydrogen, the hydrogen storage tank stores the hydrogen produced by the electrolyzer, and the DC-DC step-down converter adjusts the input electrical energy of the electrolyzer through a switch control signal.

[0008] Collaborative communication module: Utilizing a communication network as its core, it is used to receive and transmit key status information of the system in real time. The key status information includes weather conditions, load demand, and electricity purchase demand. Based on the key status information, the system's operating mode is determined, and the operating mode is used as the input condition for system operation.

[0009] The distributed optimization control module includes a photovoltaic-side controller and a hydrogen production-side controller. It is used to realize distributed economic optimization and collaborative control of the photovoltaic conversion module and the hydrogen energy storage module. It exchanges global state information through a collaborative communication module and implements rolling optimization control of the power electronic converters of the photovoltaic conversion module and the hydrogen energy storage module based on the system dynamic model. It solves the multi-objective cost function with the goal of optimizing the overall economic benefit of the system. The optimization problem is modeled as a non-convex mixed-integer nonlinear programming problem and solved using an improved mixed-integer nonlinear programming MINLP algorithm. The optimal switching control signal is output.

[0010] Furthermore, in the photovoltaic conversion module, the current-voltage characteristics of the photovoltaic array satisfy: Among them, I pv It is the output current of the photovoltaic array, n p and n sThese represent the number of photovoltaic cells connected in parallel and series, respectively. A is the coefficient of the exponential term used to fit the current-voltage characteristics of the photovoltaic array, and V... pv V represents the output voltage of the photovoltaic array. oc and I sc These are the open-circuit voltage and short-circuit current of the photovoltaic cell, φ(V) pv The symbol ) represents the functional relationship between the output current and output voltage of a photovoltaic array.

[0011] Furthermore, in the photovoltaic conversion module, the DC-DC boost converter is connected to the DC bus, and its dynamic model is established based on Kirchhoff's laws, with the following formulas: Among them, C s The capacitor represents the boost converter. It is the derivative of the photovoltaic array output voltage with respect to time, I pv It is the output current of the photovoltaic array. It is the time derivative of the current through the inductor of the boost converter, V. pv V represents the output voltage of the photovoltaic array. dc It is the DC bus voltage, L s For the inductor of the boost converter, I L S represents the current flowing through the inductor. s S is the switching control signal for the boost converter. s When = 0, the switch is open, S s When V = 1, the switch is closed, and V dc This is the DC bus voltage; the dynamic formula for the DC bus voltage is: Among them, C dc For DC bus capacitor, I a For the output current of the photovoltaic system, I b I is the input current of the DC load. g Main grid output current, I c To input current into the hydrogen production system, It is the derivative of the DC bus voltage with respect to time.

[0012] Furthermore, in the hydrogen energy storage module, the voltage characteristic formula of the electrolyzer is: Among them, V ae It is the input voltage of the electrolytic cell, I ae U is the input current to the electrolytic cell. rev For reversible battery voltage, A s T represents the area of ​​the electrolytic cell. ae n is the temperature of the electrolytic cell. ae R1 and R2 are the number of electrolytic cells connected in series, r1 and r2 are the ohmic resistance parameters related to the electrolyte, and t1, t2, t3, t eThe electrode overvoltage coefficient, φ(I) ae The sign () represents the functional relationship between the input voltage and input current of the electrolytic cell; the power formula of the electrolytic cell is: P ae =V ae I ae =φ(I ae )I ae , where P ae It is the input power of the electrolytic cell.

[0013] Furthermore, in the hydrogen energy storage module, the DC-DC buck converter is connected to the DC bus, and its dynamic model is established based on Kirchhoff's laws, as shown in the following formula: Among them, V ae It is the input voltage of the electrolytic cell, I ae The input current to the electrolytic cell, V dc This is the DC bus voltage. It is the derivative of the electrolytic cell input current with respect to time, L c For the buck converter inductor, S c S is the switching control signal for the buck converter. c When = 0, the switch is open, S c When = 1, the switch is closed; the dynamic pressure formula of the hydrogen storage tank is: in, It is the derivative of the hydrogen storage tank pressure with respect to time, P. hs For the hydrogen storage tank pressure, T o V represents the temperature of the hydrogen storage tank. hs Let R be the volume of the hydrogen storage tank, and R be the gas constant. This represents the total hydrogen production rate of the electrolyzer.

[0014] Furthermore, in the hydrogen energy storage module, the total hydrogen production rate of the electrolyzer is defined as: in, η represents the total hydrogen production rate of the electrolyzer. F Let F be the current efficiency, and I be the Faraday constant. ae For the input current of the electrolytic cell, n ae This represents the number of electrolytic cells connected in series.

[0015] Furthermore, in the collaborative communication module, the formula for determining the working mode is: Where, δ a For the operating mode of the hydrogen production system, δ g The main power grid's operating mode, δ a =1, δ g =1 corresponds to system startup, δ a =0, δ g =0 corresponds to system shutdown, P max P represents the maximum power of the photovoltaic system.L This represents the power of the DC load.

[0016] Furthermore, in the distributed optimization control module, the economic cost function of the photovoltaic-side controller is: h s (x(k),u(k))=w s1 h s1 +w s2 h s2 +w s3 h s3 +w s4 h s4 , where h s (x(k),u(k)) is the total economic cost function of the photovoltaic-side controller, x(k) is the state variable of the system at time k, u(k) is the control variable of the system at time k, and w s1 w s2 w s3 w s4 For different weighting coefficients, h s1 h s2 h s3 h s4 For the component economic costs on the photovoltaic side, the EMPC optimization problem of the photovoltaic-side controller is defined as follows: S s (k+γ)∈{0,1}, where min is the sign of the minimum value. This is the summation symbol, where γ is the step size index for the prediction time. It is the predicted state value of the system at the predicted time k+γ. It is the control prediction value of the system at the prediction time k+γ. The system outputs the predicted value at prediction time k+γ+1. It is the predicted output voltage of the photovoltaic array at time k+γ, n s V represents the number of photovoltaic cells connected in series. oc S is the open-circuit voltage. s (k+γ) is the boost converter switching control signal at the predicted time k+γ, x(k) is the system state variable at time k, f(·) is the system state transition function, g(·) is the system output function, and N p To predict the step size.

[0017] Furthermore, in the distributed optimization control module, the economic cost function of the hydrogen production side controller is: h c (x(k),u(k))=w c1 h c1 +w c2 h c2 +wc3 h c3 +w c4 h c4 +w c5 h c5 , where w c 1. w c2 w c3 w c4 w c5 For different weighting coefficients, h c (x(k), u(k)) is the total economic cost function of the hydrogen production side controller, where x(k) is the system state variable at time k, and u(k) is the system control variable at time k; the EMPC optimization problem of the hydrogen production side controller is defined as: S c (k+γ)∈{0,1}, where P hs,min P hs,max These are the minimum and maximum pressures of the hydrogen storage tank, respectively. Predicted hydrogen storage tank pressure at time k+γ, S c (k+γ) Predict the switching control signal of the buck converter at time k+γ, and find the minimum sign of the minimum value. This is the summation symbol, where γ is the step size index for the prediction time, ranging from 0 to N. p -1 indicates that the cost is accumulated over all times in the prediction time domain, γ is the step size index of the prediction time, used to traverse each time point in the prediction time domain, and N p It refers to the forecast step size, which is the forecast time domain length of the economic model forecast control (EMPC), representing the forecast of the future N. p The system state and cost at each moment It is the predicted state value of the system at the predicted time k+γ. is the control prediction value of the system at the prediction time k+γ, which is an estimate of the control variables at future times. f(·) is the system state transition function, which describes the change of the system state from the current prediction time k+γ to the next prediction time k+γ+1. is the predicted output value of the system at the prediction time k+γ+1, which is an estimate of the system output parameters at future times. g(·) is the system output function, which describes the mapping relationship between the system state and the system output.

[0018] Furthermore, in the distributed optimization control module, the improved MINLP algorithm has the following specific steps:

[0019] Initialization: Set the prediction step size N p The initial state of the system x(0) and the initial value of the optimal cost are determined to define the constraint range of each control variable and state variable.

[0020] Call the recursive search function: Set the initial search depth depth=1, and start the recursive search process;

[0021] Depth determination: Determine the current search depth (depth) and the prediction step size (N). p Size relationship;

[0022] Controlling input traversal and state prediction: If depth≤N p It iterates through all possible control inputs at the current depth, predicts the system state at the next moment based on the system dynamic model, calculates the cost of the current stage, and accumulates the total cost before the current depth.

[0023] Pruning and Optimal Update: Determine if the cumulative cost is greater than the current optimal cost. If it is, prune the current search branch and return to the control variable enumeration step. If it is not greater, update the optimal cost, record the current control variable, increment the search depth by 1, and return to the depth judgment step.

[0024] Optimal control output: If depth > N p Output the optimal control variable sequence recorded in the pruning and optimal update steps, and apply the first control variable in the optimal control variable sequence to the corresponding power electronic converter.

[0025] Compared with existing technologies, this distributed collaborative control system for photovoltaic hydrogen production DC microgrids has the following advantages:

[0026] I. This invention constructs an integrated collaborative control framework comprising a photovoltaic module, a hydrogen production module, a collaborative communication network module, and a distributed EMPC module. It replaces the traditional centralized system with a distributed control architecture. Within the distributed EMPC module, independent local controllers are designed for both the photovoltaic and hydrogen production modules. Through the collaborative communication network, global status information is exchanged in real time, enabling unified local optimization decisions and global collaborative scheduling across modules. This design effectively avoids the high communication latency, poor scalability, and single-point failure risks of traditional centralized systems. Simultaneously, through rolling optimization control, it dynamically coordinates the operating states of photovoltaic power generation and hydrogen energy storage, maintaining power supply and demand balance and DC bus voltage stability. This solves the problems of large dynamic differences and high coordination difficulty among subsystems in traditional hierarchical control, promoting the efficient coupled operation of photovoltaic power generation and hydrogen energy storage.

[0027] II. This invention introduces an improved mixed-integer nonlinear programming algorithm to efficiently solve the nonconvex optimization problem of the system, overcoming the limitations of traditional control methods that suffer from heavy computational burdens and neglect of economic efficiency and equipment losses. In the cost function design of the distributed EMPC module, multiple objectives such as power balance, equipment operating losses, maximum capacity, and economic benefits are comprehensively incorporated. Through weight adjustment, the priority of each objective is scientifically coordinated, ensuring both the basic operational stability of the system and maximizing overall economic benefits. Compared to traditional control methods that only focus on voltage and power balance, this system can dynamically adjust the hydrogen production strategy and power purchase plan under fluctuations in sunlight and load, converting excess electrical energy into hydrogen energy storage to reduce waste, while simultaneously reducing equipment switching losses, significantly improving the operating efficiency, economy, and disturbance rejection capability of the DC microgrid.

[0028] Other advantages, objectives and features of the invention will be set forth in part in the description which follows, and in part will be apparent to those skilled in the art from the following examination or study, or may be learned from the practice of the invention. Attached Figure Description

[0029] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the accompanying drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are merely some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without any creative effort.

[0030] Figure 1 This is a schematic diagram of a distributed collaborative control system for a photovoltaic hydrogen production DC microgrid.

[0031] Figure 2 This is a structural diagram of a distributed collaborative control system for a photovoltaic hydrogen production DC microgrid.

[0032] Figure 3 This is a flowchart of the MINLP algorithm for a distributed collaborative control system of a photovoltaic hydrogen production DC microgrid.

[0033] Figure 4 This is a voltage and current diagram of the photovoltaic and electrolyzer cells in a distributed collaborative control system for a photovoltaic hydrogen production DC microgrid.

[0034] Figure 5 This is a diagram showing the DC bus voltage and power of each subsystem in a distributed collaborative control system for a photovoltaic hydrogen production DC microgrid.

[0035] Figure 6 This is a diagram showing the pressure change of the hydrogen storage tank and the photovoltaic absorption of a distributed collaborative control system for a photovoltaic hydrogen production DC microgrid. Detailed Implementation

[0036] To further illustrate the technical means and effects of the present invention in achieving its intended purpose, the following detailed description of the specific implementation methods, structures, features, and effects of the present invention, in conjunction with the accompanying drawings and preferred embodiments, is provided below.

[0037] Example 1:

[0038] Establish a grid-connected microgrid model comprising a photovoltaic array, DC loads, an electrolyzer, and a hydrogen storage tank. Assume the rated power of the photovoltaic array is 36kW, the rated power of the electrolyzer is 10kW, and the upper limit of the purchased power is 10kW.

[0039] To simulate typical solar irradiance variations, the irradiance was maintained at 1000 W / m² for 0–0.1 s. 2 It then rose to 1100W / m 2 It remains stable for 0.2–0.3 seconds, then gradually decreases after 0.3 seconds. The load will suddenly increase from 30kW to 45kW at 0.3 seconds, increasing the pressure on the system's supply and demand regulation.

[0040] Figure 4 The dynamic changes in voltage and current of the photovoltaic system and the electrolytic cell were demonstrated separately. The photovoltaic system voltage remained relatively stable at approximately 180–190V, exhibiting strong stability with only minor fluctuations during periods of increased sunlight (0.1–0.3s) and load jumps (0.3s). This indicates that the controller effectively maintained voltage stability, which is beneficial for maximum power point tracking and stable output. The photovoltaic system current responded significantly to changes in sunlight and load: initially 195A, gradually increasing from 0.1–0.2s with increased sunlight, reaching a peak of 217A from 0.2–0.3s. After the load suddenly increased to 45kW at 0.3s, the system redistributed power, and the photovoltaic system current rapidly decreased and tended towards a new steady-state value, demonstrating the system's ability to quickly adapt to disturbances. The electrolytic cell voltage gradually increased from 0 to 0.3s, reflecting the increased workload of the electrolytic cell as surplus power increased; after 0.3s, due to the sudden load increase, the system prioritized power supply to the load, and the electrolytic cell stopped operating. The current in the electrolyzer rises from 100A to 150A within 0 to 0.3 seconds, indicating that the electrolyzer is in a high-load hydrogen production state during this stage; after the load jumps, the current drops rapidly to zero.

[0041] Figure 5The DC bus voltage and power of each subsystem are displayed. In the initial stage (0–0.3s), although the bus voltage fluctuated slightly, it remained stable between 301 and 302V, indicating a good steady state. At t = 0.3s, the load suddenly increased from 30kW to 45kW, causing a significant disturbance and a short-term drop in bus voltage. At this time, the system achieved dynamic balance by rapidly adjusting photovoltaic output, purchased power, and electrolytic cell power: the electrolytic cell power quickly dropped to zero, the main grid output power increased to compensate for the increased load, and the photovoltaic power remained close to its maximum power point. Thanks to the effective response of the EMPC control strategy, the bus voltage quickly recovered and stabilized at approximately 300V within the 0.3–0.5s range, with significantly reduced fluctuations. The power curve remained smooth and without abrupt changes throughout the process, fully verifying the dynamic adjustment capability and voltage stability performance of the control strategy under load abrupt changes.

[0042] Figure 6 This simulation illustrates the pressure changes in a hydrogen storage tank within a microgrid. Initially, the tank pressure is 1000 Pa. During the first 0-0.3 seconds, due to a low system load (30 kW) and sufficient photovoltaic output, the EMPC controller detects excess power and keeps the electrolyzer running, causing the tank pressure to rise continuously. At 0.3 seconds, the system load suddenly increases to 45 kW, exceeding the maximum photovoltaic output. At this point, the EMPC controller automatically determines a power shortage, quickly stops the electrolyzer, and prioritizes power supply to the load, thus halting hydrogen production. The tank pressure stops rising and stabilizes at approximately 3850 Pa. This figure verifies the rapid response of the electrolyzer system and the effectiveness of the hydrogen storage tank, while also demonstrating the dynamic adaptability of the EMPC in power allocation: producing hydrogen when energy is abundant and prioritizing the load during periods of stress, ensuring system stability and maximizing energy efficiency. Figure 6 This section describes the relationship between the actual utilization of photovoltaic (PV) power in the system and the load demand from the perspective of PV integration. The square lines represent the remaining power (P2) after deducting the power consumed by the electrolyzer from the total PV power generated and purchased. pv +P g -P ae This represents the system's capacity to absorb photovoltaic power after meeting hydrogen production needs, which can then be used for other loads. The circled line represents the system's load power P. L The triangle represents the error P between the two. error As shown in the graph, the system experiences a relatively low load in the first 0.3 seconds, allowing for effective absorption of photovoltaic (PV) power. However, after a surge in load, the system adjusts its power purchase to maintain overall supply-demand balance. This graph reflects the dynamic process of PV energy utilization and absorption within the system, effectively converting excess electrical energy into chemical energy in the form of hydrogen for energy storage. Figure 2 As shown.

[0043] Example 2:

[0044] A specific implementation of a distributed collaborative control system for photovoltaic hydrogen production DC microgrids.

[0045] This embodiment uses a microgrid system of "photovoltaic power supply - hydrogen energy storage - DC load guarantee" in a small manufacturing park as an application scenario. The park requires a stable DC load supply for daily operation, with load power fluctuating according to production periods, ranging from 50kW to 120kW per day. Simultaneously, the park has a fuel cell backup power system that requires regular hydrogen replenishment. Therefore, photovoltaic hydrogen production is needed to meet part of the hydrogen supply demand, reducing the cost of purchasing hydrogen and dependence on the municipal power grid. The core equipment configuration of the system is strictly designed according to this invention, specifically including, for example... Figure 3 As shown:

[0046] Photovoltaic conversion module: 150kW photovoltaic array, paired with a 200kW DC-DC boost converter;

[0047] Hydrogen energy storage module: 1 x 80kW alkaline electrolyzer, 1 x 100kW DC-DC step-down converter, 1 x 35m³ capacity unit 3 High-pressure hydrogen storage tank;

[0048] Collaborative communication module: Based on an industrial wireless communication network, equipped with a light sensor, an ambient temperature sensor, a DC load power monitor, a hydrogen storage tank pressure sensor, and a power grid purchase price acquisition terminal;

[0049] Distributed optimization control module: 2 industrial-grade embedded controllers are used to achieve collaborative optimization of the two major modules.

[0050] The system automatically starts at 7:00 AM every day. The distributed optimization control module first completes the initialization operation: First, it sets the prediction step size, which is set to 48 based on the park's production rhythm and photovoltaic power variation. Second, it enters the system's initial state, including the initial operating state of the photovoltaic array under low light conditions in the early morning, the initial pressure of the hydrogen storage tank after the previous day's shutdown, and the DC load power during the park's off-peak production period in the morning. Third, it clarifies the constraints of control variables and state variables. The switching control signals of the photovoltaic-side DC boost converter and the hydrogen production-side DC buck converter can only switch between "open" and "closed" states. The output voltage of the photovoltaic array must not exceed the maximum open-circuit voltage determined by the number of batteries connected in series. The pressure of the hydrogen storage tank must always be maintained between the minimum safe pressure and the maximum rated pressure. Fourth, it sets the initial value of the optimal cost. Referring to the optimal operating data of the microgrid in the park over the past 3 months, the initial optimal cost is set to 800 yuan / 4 hours.

[0051] The collaborative communication module collects key system status information in real time at a frequency of 1 second: it obtains the current light intensity of the park through a light sensor, the current ambient temperature (23℃) through an ambient temperature sensor, the real-time load demand of the park through a DC load power monitor, and the current municipal grid electricity purchase price through a grid purchase price acquisition terminal. Based on the above information, the collaborative communication module first estimates the current maximum power of the photovoltaic array (approximately 70kW, considering both light intensity and photovoltaic array characteristics), and then compares the relationship between the maximum photovoltaic power and the load power: the current maximum photovoltaic power (70kW) is greater than the load power (50kW). According to the working mode determination rules, the collaborative communication module finally determines the system working mode as "hydrogen production system started, main grid shut down," and uses this working mode as the core input condition, synchronously transmitting it to the distributed optimization control module, photovoltaic conversion module, and hydrogen energy storage module.

[0052] Photovoltaic array under current illumination (300W / m) 2 Under ambient temperature (23℃), the system continuously converts solar energy into DC power, and its output current and voltage characteristics are affected by environmental conditions. One end of the DC-DC boost converter is connected to the photovoltaic array, and the other end is connected to the park's DC bus, receiving initial switching control signals from the distributed optimization control module. When the output voltage of the photovoltaic array briefly drops due to cloud cover, the DC-DC boost converter quickly stabilizes the voltage output to the DC bus at 750V by switching the control signals and adjusting the operating states of the internal capacitors and inductors based on circuit rules. This ensures stable power supply to the park's 50kW DC load, meeting the operational needs of production equipment; and it also transmits excess power generated by the photovoltaic array to the DC bus, providing power support for the hydrogen energy storage module.

[0053] Since the collaborative communication module has determined that the "hydrogen production system has started," the DC-DC step-down converter in the hydrogen energy storage module first receives the switching control signal from the distributed optimization control module, reducing the 750V voltage of the DC bus to the operating voltage required by the electrolyzer, thus precisely regulating the electrical energy input to the electrolyzer. After receiving electrical energy, the electrolyzer decomposes deionized water into hydrogen and oxygen through internal electrode reactions. The oxygen is directly discharged through the park's waste gas treatment pipeline, while the hydrogen is transported to a high-pressure hydrogen storage tank for storage via a dedicated pipeline. The hydrogen storage tank monitors its internal pressure in real time using a pressure sensor. The initial pressure is an intermediate value, and as hydrogen continues to be added, the pressure rises slowly at a stable rate. When the light intensity reaches 750W / m² after 10:00 AM... 2 The maximum power of the photovoltaic array is increased to 140kW, while the load power is still 50kW. The excess power increases to 90kW. Under the regulation of the control signal, the DC-DC step-down converter increases the input power of the electrolyzer to 90kW, and the hydrogen production rate accelerates accordingly. The pressure rise rate of the hydrogen storage tank also increases accordingly.

[0054] The photovoltaic-side controller and the hydrogen production-side controller exchange global status information every 5 minutes through a collaborative communication module. This exchange includes real-time output power of the photovoltaic array, DC bus voltage, real-time DC load power, current pressure of the hydrogen storage tank, input power to the electrolyzer, and current grid electricity price. With the goal of achieving optimal overall system economic efficiency, both controllers implement rolling optimization control of the power electronic converters in the photovoltaic conversion module and the hydrogen storage module based on the system's real-time operating status. The photovoltaic-side controller focuses on calculating the operating losses of the photovoltaic array and the energy consumption of the boost converter, determining the switching control sequence of the boost converter through optimization. The hydrogen production-side controller focuses on calculating the energy consumption of the electrolyzer and the pressure maintenance cost of the hydrogen storage tank, determining the switching control sequence of the buck converter through optimization. During the optimization process, the system transforms the overall optimization problem into a non-convex mixed-integer nonlinear programming problem, which is solved using an improved mixed-integer nonlinear programming algorithm: First, it confirms that initialization has been completed; then, it calls a recursive search function, starting from the initial search depth, and determines whether the current depth is less than or equal to the prediction step size; if so, it iterates through all possible switching control signals at the current depth, predicts the state at the next moment based on the system's operating rules, calculates the economic cost of the current stage and accumulates the total cost; if the accumulated cost is greater than the current optimal cost, it discards the control signal branch; if it is less than or equal to, it updates the optimal cost and records the current control signal, increases the search depth by 1 and continues to determine the next step size; if the search depth exceeds the prediction step size, it outputs the recorded optimal control signal sequence, applies the first control signal in the sequence to the corresponding DC-DC boost converter or DC-DC buck converter, and ensures that the system operating cost is minimized during the current period.

[0055] At 3:00 PM, a sudden, brief thunderstorm occurred in the park, causing the sunlight intensity to drop sharply to 150W / m². 2 After the collaborative communication module collects this change in real time, it re-estimates the maximum power of the photovoltaic array. At this point, the maximum photovoltaic power is less than the load power. The collaborative communication module immediately re-determines the system operating mode: the hydrogen production system is shut down, the main grid is started, and the new mode is transmitted to each module. After receiving the new mode, the distributed optimization control module restarts the improved mixed-integer nonlinear programming algorithm and adjusts the control signals: the photovoltaic-side boost converter remains "closed" to maximize photovoltaic output; the hydrogen production-side buck converter switches to "open" to stop hydrogen production; at the same time, it optimizes the main grid power supply to ensure that the electricity purchase cost is minimized while meeting the 100kW load demand, until the thunderstorm ends and the sunlight intensity recovers, at which point the system switches back to the "photovoltaic power supply + hydrogen production and storage" mode. Figure 1 As shown.

[0056] The above description is merely a preferred embodiment of the present invention and is not intended to limit the present invention in any way. Although the present invention has been disclosed above with reference to preferred embodiments, it is not intended to limit the present invention. Any person skilled in the art can make some modifications or alterations to the above-disclosed technical content to create equivalent embodiments without departing from the scope of the present invention. Any simple modifications, equivalent changes and alterations made to the above embodiments based on the technical essence of the present invention without departing from the scope of the present invention shall still fall within the scope of the present invention.

Claims

1. A distributed cooperative control system for a photovoltaic hydrogen generation direct current microgrid, characterized in that, The system comprises: A photovoltaic conversion module: composed of a photovoltaic array and a DC boost converter, used for converting solar energy into electrical energy, the operating state parameters of the photovoltaic array are affected by environmental conditions, and the DC boost converter adjusts operation through a switching control signal; A hydrogen energy storage module: composed of an electrolytic cell, a DC step-down converter and a hydrogen storage tank, used for energy conversion and storage, the electrolytic cell receives electrical energy to produce hydrogen, the hydrogen storage tank stores hydrogen produced by the electrolytic cell, and the DC step-down converter adjusts the input power of the electrolytic cell through a switching control signal; A collaborative communication module: using a communication network as the core, used for real-time reception and transmission of key state information of the system, the key state information includes weather conditions, load demand and electricity purchase demand, based on the key state information to determine the system working mode, the working mode as the input condition of the system operation; A distributed optimization control module: including a photovoltaic side controller and a hydrogen production side controller, used for realizing distributed economic optimization and collaborative control of the photovoltaic conversion module and the hydrogen energy storage module, exchanging global state information through the collaborative communication module, based on the system dynamic model to implement rolling optimization control of the power electronic converters of the photovoltaic conversion module and the hydrogen energy storage module, using the overall economic benefit optimization of the system as the target to solve the multi-objective cost function, the optimization problem is modeled as a non-convex mixed integer nonlinear programming problem, and an improved mixed integer nonlinear programming MINLP algorithm is used to solve, and the optimal switching control signal is output.

2. The distributed cooperative control system for photovoltaic hydrogen generation DC microgrid according to claim 1, characterized in that, The current-voltage characteristic of the photovoltaic array in the photovoltaic conversion module satisfies: wherein, I pv is the output current of the photovoltaic array, n p and n s are the parallel quantity and series quantity of the photovoltaic cells respectively, A is the exponential term coefficient for fitting the current-voltage characteristic of the photovoltaic array, V pv represents the output voltage of the photovoltaic array, V oc and I sc are the open circuit voltage and short circuit current of the photovoltaic cell respectively, and φ(V pv ) represents the symbol of the function relationship between the output current and the output voltage of the photovoltaic array. 3.The distributed collaborative control system for photovoltaic hydrogen generation DC micro-grid of claim 1, wherein, The photovoltaic conversion module, DC boost converter access DC bus, its dynamic model is established according to Kirchhoff's law, formula is respectively: Wherein, C s Represent the capacitor of boost converter, Is the derivative of photovoltaic array output voltage to time, I pv Is the output current of photovoltaic array, Is the derivative of current through boost converter inductance to time, V pv Photovoltaic array output voltage, V dc Is the DC bus voltage, L s For boost converter inductance, I L Represent the current through inductance, S s For boost converter switch control signal, S s S s = 0 when the switch is off, S dc = 1 when the switch is closed, V dc Is the DC bus voltage; the dynamic formula of the DC bus voltage is: Wherein, C a For DC bus capacitor, I b For photovoltaic system output current, I g For DC load input current, I c For main grid output current, Is the derivative of DC bus voltage to time. 4.The distributed collaborative control system for photovoltaic hydrogen generation DC micro-grid of claim 1, wherein, The voltage characteristic formula of the electrolytic cell in the hydrogen energy storage module is: Wherein, V ae is the input voltage of the electrolytic cell, I ae is the input current of the electrolytic cell, U rev is the reversible battery voltage, A s is the electrolytic cell area, T ae is the electrolytic cell temperature, n ae is the number of electrolytic cell series, r1 and r2 are ohmic resistance parameters related to the electrolyte, t1, t2, t3 and t e are electrode overvoltage coefficients, and φ(I ae ) represents the symbol of the function relationship between the input voltage and the input current of the electrolytic cell; the power formula of the electrolytic cell is: P ae = V ae I ae = φ(I ae )I ae , wherein P ae is the input power of the electrolytic cell.

5. The distributed cooperative control system for photovoltaic hydrogen generation DC microgrid according to claim 1, characterized in that, The hydrogen energy storage module, DC voltage converter is connected with DC bus, its dynamic model is established according to Kirchhoff's law, formula is: Wherein, V ae Is the input voltage of electrolytic cell, I ae For electrolytic cell input current, V dc For DC bus voltage, Is the derivative of electrolytic cell input current to time, L c For voltage converter inductance, S c For voltage converter switch control signal, S c = 0 when the switch is off, S c = 1 when the switch is closed;The pressure dynamic formula of the hydrogen storage tank is: Wherein, Is the derivative of hydrogen storage tank pressure to time, P hs For hydrogen storage tank pressure, T o For hydrogen storage tank temperature, V hs For hydrogen storage tank volume, R is gas constant, For the total hydrogen production rate of electrolytic cell.

6. The distributed cooperative control system for photovoltaic hydrogen generation DC microgrid according to claim 5, characterized in that, The total hydrogen production rate of the electrolytic cell is defined as: wherein, is the total hydrogen production rate of the electrolytic cell, η F is the current efficiency, F is the Faraday constant, I ae is the input current of the electrolytic cell, n ae is the number of series of the electrolytic cell. 7.The distributed cooperative control system for photovoltaic hydrogen generation DC microgrid of claim 1, wherein, The determination formula of the working mode in the cooperative communication module is: Wherein, δ a is the working mode of the hydrogen production system, δ g is the working mode of the main power grid, δ a =1, δ g =1 corresponds to system startup, δ a =0, δ g =0 corresponds to system shutdown, P max is the maximum power of the photovoltaic system, P L is the DC load power. 8.The distributed collaborative control system for photovoltaic hydrogen generation DC micro-grid of claim 1, wherein, The economic cost function of the photovoltaic side controller in the distributed optimization control module is: s (x(k),u(k))=w s1 h s1 +w s2 h s2 +w s3 h s3 +w s4 h s4 wherein h s (x(k),u(k)) is the total economic cost function of the photovoltaic side controller, x(k) is the state variable of the system at time k, u(k) is the control variable of the system at time k, w s1 , w s2 , w s3 , w s4 are different weight coefficients, h s1 , h s2 , h s3 , h s4 are the sub-economic costs of the photovoltaic side, and the EMPC optimization problem of the photovoltaic side controller is defined as: S s (k+γ)∈{0,1},wherein min denotes the minimum value symbol, is the summation symbol, wherein γ is the step index of the prediction time, is the state prediction value of the system at the prediction time k+γ, is the control prediction value of the system at the prediction time k+γ, is the output prediction value of the system at the prediction time k+γ+1, is the photovoltaic array output voltage prediction value at the prediction time k+γ, n s is the number of series of photovoltaic cells, V oc is the open circuit voltage, S s (k+γ) is the boost converter switch control signal at the prediction time k+γ, x(k) is the state variable of the system at time k, f(·) is the system state transition function, g(·) is the system output function, N p is the prediction step. 9.The distributed collaborative control system for photovoltaic hydrogen generation DC micro-grid of claim 1, wherein, The economic cost function of the hydrogen production side controller in the distributed optimization control module is h c (x(k),u(k))=w c1 h c1 +w c2 h c2 +w c3 h c3 +w c4 h c4 +w c5 h c5 , wherein w c 1, w c2 , w c3 , w c4 , w c5 are different weight coefficients, h c (x(k),u(k)) is the total economic cost function of the hydrogen production side controller, x(k) is the state variable of the system at time k, and u(k) is the control variable of the system at time k; the EMPC optimization problem of the hydrogen production side controller is defined as S c (k+γ)∈{0,1}, wherein P hs,min , P hs,max are the minimum pressure and the maximum pressure of the hydrogen storage tank, is the predicted value of the hydrogen storage tank pressure at prediction time k+γ, S c (k+γ) is the pressure reducer switch control signal at prediction time k+γ, min is the minimum value symbol, is the summation symbol, wherein γ is the step index of the prediction time, from 0 to N p -1, indicating that the costs of all times in the prediction time domain are accumulated, γ is the step index of the prediction time, used to traverse each time in the prediction time domain, N p is the prediction step, that is, the prediction time domain length of the economic model prediction control (EMPC), representing the system state and cost of the future N p times, is the state prediction value of the system at prediction time k+γ, is the control prediction value of the system at prediction time k+γ, which is an estimate of the control variable at future times, f(·) is a system state transition function, describing the change rule of the system state from the current prediction time k+γ to the next prediction time k+γ+1, is the output prediction value of the system at prediction time k+γ+1, which is an estimate of the system output parameter at future times, and g(·) is a system output function, describing the mapping relationship between the system state and the system output.

10. The distributed cooperative control system for photovoltaic hydrogen generation DC microgrid according to claim 1, characterized in that, In the distributed optimization control module, the specific steps of the improved MINLP algorithm are: Initialization: Set the prediction step N p , the initial state of the system x(0), the initial value of the optimal cost, and determine the constraint range of each control variable and state variable; Call recursive search function: set initial search depth depth = 1, start recursive search process; Depth judgment: judging the size relationship between the current search depth depth and the predicted step length N p ; Control input enumeration and state prediction: if depth≤N p , enumerate all possible control inputs at the current depth, predict the system state at the next time step based on the system dynamic model, compute the cost of the current stage, and accumulate the total cost before the current depth; Pruning and optimal update: judge whether the cumulative cost is greater than the current optimal cost, if greater, prune the current search branch, and return to the control variable enumeration step; If not greater than, update the optimal cost, record the current control variable, increase the search depth depth by 1, and return to the depth judgment step; Optimal control output: if depth > N p , output the pruning and the optimal control variable sequence recorded in the optimal update step, and apply the first control variable in the optimal control variable sequence to the corresponding power electronic converter.

Images