Collaborative hierarchical control system for PEM-ALK hybrid hydrogen production DC microgrid system

Abstract

The present invention discloses a collaborative hierarchical control system suitable for a PEM-ALK hybrid hydrogen production DC microgrid system. The system includes a power distribution control layer, a strategy management control layer, and a device control layer; the power distribution control layer is used to achieve power balance of the DC microgrid system using the wavelet decomposition method; the strategy management control layer is used to switch the control strategy of the battery and the working state of the electrolyzer, thereby improving the system's collaborative working performance while ensuring bus voltage stability; the device control layer is used to determine the unique expected operating voltage U corresponding to the maximum hydrogen production efficiency of the ALK electrolyzer and the PEM electrolyzer under different operating power conditions based on the current-hydrogen production efficiency characteristic curve. ref , then establish the fractional-order PI of the Buck converter λ The control system makes the actual working voltage of the electrolytic cell consistent with the expected working voltage U ref The present invention achieves power balance of the system, reduces the number of times the battery is charged and discharged, and effectively improves the hydrogen production efficiency and stability of the electrolyzer.

Description

Technical Field

[0001] The present invention relates to the technical field of wind power hydrogen production operation control, and specifically to a collaborative hierarchical control system suitable for a PEM-ALK hybrid hydrogen production DC microgrid system. Background Art

[0002] In modern power systems, the proportion of clean energy continues to rise. However, renewable energy sources such as solar and wind power are highly volatile and random, and their direct integration into the grid can significantly impact the safe and stable operation of the grid. DC microgrids significantly improve energy efficiency by reducing energy losses during power conversion, effectively alleviating the challenges of integrating renewable energy into the grid.

[0003] Compared to AC power, DC microgrid stability assessment is more straightforward, focusing primarily on the bus voltage. DC loads can be connected directly to the DC bus, eliminating the need for complex DC / AC conversion. This not only simplifies the system structure and avoids complex issues such as frequency collapse and phase shift, but also significantly reduces losses during energy conversion, further improving system efficiency. In terms of control, DC microgrids offer reduced control difficulty and failure rates compared to other grid types, significantly facilitating operations and maintenance.

[0004] On the other hand, hydrogen, as an efficient, clean, and sustainable energy source, demonstrates tremendous potential for application in a wide range of fields. In-depth research and active promotion of hydrogen energy technology will strongly promote the vigorous development of related industries. It will complement green energy technologies such as DC microgrids, jointly promoting improved energy efficiency, reducing environmental pollution, and providing solid support for sustainable development. Electrolyzer hydrogen production technology, a key technology bridging the gap between electricity and hydrogen—the future of clean energy—is invaluable for building a sustainable, low-carbon energy system due to its efficient and clean energy conversion characteristics.

[0005] The organic combination of wind power generation, hybrid energy storage, electrolyzer hydrogen production, and DC microgrids is expected to build a more efficient, clean, and sustainable energy system. However, there are the following technical problems:

[0006] First, due to the complexity and nonlinearity of battery charging and discharging, traditional power allocation strategies may not be sufficient to meet the energy requirements of various energy storage systems.

[0007] Second, the battery is charged and discharged too many times, resulting in overcharging and over-discharging problems, which shortens the battery life;

[0008] Third, the dynamic response speed of the electrolyzer is slow and the energy conversion efficiency is unstable. Summary of the Invention

[0009] In response to the shortcomings of the existing technology, the present invention proposes a collaborative hierarchical control system suitable for a PEM-ALK hybrid hydrogen production DC microgrid system. On the basis of ensuring the stability of the DC microgrid bus voltage, it fully absorbs the fluctuating electric energy generated by the wind turbine to achieve power balance of the system. It not only reduces the number of battery charge and discharge times and prevents battery overcharge and over-discharge, but also effectively improves the hydrogen production efficiency and stability of the electrolyzer.

[0010] To achieve the above objectives, the present invention designs a collaborative hierarchical control system suitable for a PEM-ALK hybrid hydrogen production DC microgrid system, which is used to accurately control the PEM-ALK hybrid hydrogen production DC microgrid system. The unique features of the control system are: it includes a power distribution control layer, a policy management control layer, and a device control layer;

[0011] The DC microgrid system includes a wind turbine power generation module, a flywheel energy storage module, a battery energy storage module, an electrolyzer hydrogen production module, and an AC power grid module, each of which is connected to a DC bus. The electrolyzer hydrogen production module includes an ALK electrolyzer hydrogen production module and a PEM electrolyzer hydrogen production module connected in parallel. The ALK electrolyzer and the PEM electrolyzer are connected to the DC bus through different Buck converters.

[0012] The power distribution control layer is used to collect real-time wind speed data to obtain the wind turbine power generation power P w , in the absence of Internet communication, the wind turbine power generation P is converted into w Decomposed into a linear combination of multiple wavelet basis functions of different frequencies, reconstructed into the grid-connected power P bing , flywheel power P f and battery power P bat , under the premise of ensuring the stability of the DC bus voltage, the power balance of the DC microgrid system is achieved; at the same time, the allocated power signals are transmitted to the strategy management control layer and the device control layer respectively;

[0013] The strategy management control layer is used to collect the battery power P in real time. bat , battery voltage U bat , battery current I bat , and battery SOC, and switch the battery control strategy and the working state of the electrolyzer according to the battery SOC range. The battery switches between constant voltage control mode and constant power control mode, and changes the working power of the ALK electrolyzer and PEM electrolyzer at the same time; the working power of the ALK electrolyzer and PEM electrolyzer, the battery voltage U bat signal, and battery current I bat The signal is transmitted to the device control layer;

[0014] The device control layer is used to collect the DC microgrid bus voltage U in real time.dc , the power values ​​after distribution, the working power of the electrolytic cell, the battery voltage U bat , and battery current I bat According to the mathematical models of ALK electrolyzer and PEM electrolyzer, the current-hydrogen production efficiency characteristic curves of ALK electrolyzer and PEM electrolyzer are simulated, and then the only expected working voltage U corresponding to the maximum hydrogen production efficiency of ALK electrolyzer and PEM electrolyzer under different working power conditions is determined according to the current-hydrogen production efficiency characteristic curves. ref , establish the fractional-order PI of Buck converter λ The control system uses fractional-order PI λ The control system is connected with the Black Kite algorithm and Buck converter, and the feedback objective function is used as the evaluation index to optimize the control parameters so that the actual working voltage U ALK , Actual working voltage of PEM electrolyzer U PEM and the expected operating voltage U ref The difference between them is minimized, achieving the goal of stabilizing the ALK electrolyzer and PEM electrolyzer at the maximum efficiency point.

[0015] Furthermore, in the DC microgrid system, the wind turbine power generation module includes a wind turbine system and a converter, and the wind turbine system is connected to the DC bus through an AC / DC converter; the flywheel energy storage module includes a flywheel energy storage and a converter, and the flywheel energy storage is connected to the DC bus through a rectifier converter; the battery energy storage module includes a battery energy storage and a converter, and the battery energy storage is connected to the DC bus through a DC / DC converter; the AC power grid module includes an AC power grid and a converter, and the AC power grid is connected to the DC bus through an AC / DC converter.

[0016] Furthermore, in the power distribution control layer, the wind turbine power generation power P w The equation is

[0017]

[0018] Where,

[0019] P w is the wind turbine power generation power,

[0020] ρ is the air density,

[0021] V is the real-time wind speed,

[0022] R is the radius of the fan blade,

[0023] C p is the wind energy utilization coefficient, which is a function of the pitch angle β and the blade tip rotation ratio λ.

[0024] Furthermore, in the power distribution control layer, wavelet decomposition decomposes the original signal into a low-frequency signal A through low-pass filter and high-pass filter. J and a series of high frequency signals [D1, D2, ..., D J ], and the orthogonal and approximately symmetrical Symlets4 are used to decompose and reconstruct the signal. When the signal is decomposed to the third layer, the third layer signal can be expressed as:

[0025] P w =D1+D2+D3+A3

[0026] Where,

[0027] P w is the fan power,

[0028] D1 is a high frequency signal.

[0029] D2 is a high frequency signal,

[0030] D3 is a high frequency signal.

[0031] A3 is a low-frequency signal;

[0032] When signal D2 meets the grid connection requirements, the grid connection power P bing = D2+D3+A3, flywheel power P f and battery power P bat The sum is D1.

[0033] Furthermore, in the power distribution control layer, wavelet quadratic decomposition is used to reconstruct the D1 signal, where low-frequency power is allocated to the battery energy storage module and high-frequency power is allocated to the flywheel energy storage module. The power allocated to the battery energy storage module and the flywheel energy storage module meets the following conditions:

[0034]

[0035] Where,

[0036] P bat is the power allocated to the battery,

[0037] P bat_max is the maximum capacity of the battery,

[0038] P f is the power distributed to the flywheel,

[0039] P f_max is the maximum capacity of the flywheel.

[0040] Furthermore, in the strategy management control layer, the specific control strategy for the battery, ALK electrolyzer, and PEM electrolyzer is:

[0041] When the battery SOC is greater than 85%: If P bat >0, stop charging the battery; if P bat <0, control the battery to be in a constant voltage state; adjust the ALK electrolyzer and PEM electrolyzer at the rated working power P e Run down;

[0042] When the battery SOC is greater than 80% and less than 85%, control the battery to be in a constant voltage state, adjust the ALK electrolyzer and PEM electrolyzer to the rated working power P e Run down;

[0043] When the battery SOC is greater than 30% and less than 80%, the battery is controlled to be in a constant power state, and the ALK electrolyzer and PEM electrolyzer are adjusted to operate at the rated power P e Run down;

[0044] When the battery SOC is greater than 15% and less than 30%, the battery is controlled to be in a constant power state and the working power of the ALK electrolyzer is adjusted to 0.2P e , adjust the working power of PEM electrolyzer to 0.1P e ;

[0045] When the battery SOC is less than 15%: If P bat <0, the battery stops working; if P bat >0, the battery is controlled to be in constant power state; both the ALK electrolyzer and the PEM electrolyzer are not working.

[0046] Furthermore, in the equipment control layer, the mathematical model of the ALK electrolyzer is

[0047]

[0048] Where,

[0049] U ALK is the actual working voltage of ALK electrolyzer,

[0050] U rev is the reversible voltage,

[0051] U ohm is the resistance overvoltage caused by ohmic polarization,

[0052] U con is the concentration polarization overvoltage,

[0053] T is the working temperature of the electrolytic cell,

[0054] I is the current of the hydrogen production unit,

[0055] A ele is the cathode plate area,

[0056] r1, r2, s1, s2, s3, t1, t2, t3, α are adjustment coefficients;

[0057] The mathematical model of the PEM electrolyzer is:

[0058]

[0059] Where,

[0060] U PEM is the actual working voltage of the PEM electrolyzer,

[0061] U ocv is the open circuit voltage, which is also the minimum theoretical voltage of the PEM electrolytic cell.

[0062] U ohm is the ohmic overpotential due to the cell resistance,

[0063] U act is the overpotential due to the electrochemical reaction,

[0064] U act,a is the activation overpotential of the anode,

[0065] U act,c is the activation overpotential of the cathode.

[0066] Furthermore, in the equipment control layer, the hydrogen production efficiency η of both the ALK electrolyzer and the PEM electrolyzer can be expressed as:

[0067]

[0068] Where,

[0069] Q h is the chemical heat energy of hydrogen,

[0070] Q power is the electrical energy absorbed by the electrolytic cell,

[0071] Q heat It is the heat energy compensated to the electrolytic cell by the external heat source.

[0072] Furthermore, the fractional-order PI λ The transfer function expression of the control system is

[0073]

[0074] The fractional PI λ The output signal expression of the control system is

[0075] u(t)=K p e(t)+K I D -λe(t)

[0076] Where,

[0077] G(s) is the fractional-order PI λ The transfer function of the control system,

[0078] K P is the scaling factor optimized using the Black Kite algorithm,

[0079] K I is the integral coefficient optimized using the Black Kite algorithm,

[0080] λ is the control parameter to be optimized using the Black Kite algorithm,

[0081] u(t) is fractional-order PI λ The output signal of the control system,

[0082] e(t) is the actual working voltage U of the Buck circuit at time t out and the expected operating voltage U ref Comparing the error signal obtained,

[0083] D is fractional-order PI λ The control system obtains the control signal.

[0084] Furthermore, in the device control layer, the minimum voltage deviation and the shortest adjustment time are two evaluation indicators of the objective function, and the objective function is

[0085]

[0086] Where,

[0087] F is the objective function. The smaller the value of F, the better the solution.

[0088] a is the weight coefficient, which is used to measure the weight of the system dynamic performance.

[0089] b is the weight coefficient, which is used to measure the weight between the static performance of the system.

[0090] ITAE is an evaluation index of the system dynamic performance.

[0091] ISE is an evaluation index of the system static performance.

[0092] U ref is the desired operating voltage of the electrolyzer,

[0093] U out (t) is the actual working voltage of the electrolytic cell at time t,

[0094] T is the sampling time of the simulation.

[0095] The advantages of the present invention are:

[0096] 1. The present invention divides the control of the entire DC microgrid system into three levels: power distribution control layer, policy management control layer, and device control layer. The power distribution control layer and policy management control layer implement hierarchical control through the power signals in the system, achieving power balance of the system while ensuring the stability of the DC bus voltage. The policy management control layer and device control layer regulate and control the voltage and current signals of each module, effectively improving the operating efficiency and stability of the electrolyzer, while reducing the number of battery charge and discharge times and preventing battery overcharge and over-discharge. This regulation mechanism based on power signals and voltage and current signals is conducive to the overall optimization and fine regulation of the system.

[0097] 2. The power distribution control layer in the present invention uses the wavelet decomposition method to ensure the stability of the DC bus voltage and convert the wind turbine power P w Decomposed into a linear combination of multiple wavelet basis functions of different frequencies, reconstructed into the grid-connected power P bing , flywheel power P f and battery power P bat , achieving power balance of DC microgrid system;

[0098] 3. The strategy management control layer in the present invention controls the battery SOC to be within the specified range, and switches the battery control strategy and the working state of the electrolyzer according to the battery SOC. This not only reduces the number of charge and discharge cycles of the battery, significantly extending the battery life, but also ensures that the electrolyzer and battery module can work together efficiently according to the preset control strategy;

[0099] 4. The device control layer in the present invention establishes the fractional-order PI of the Buck converter λ Control system, using the Black Kite algorithm to optimize fractional-order PI λ The control system's control parameters improve the efficiency and stability of the hybrid hydrogen production system, enhance the overall operating performance of the system, and meet the growing demand for clean energy;

[0100] 5. For the entire system, the present invention does not require the configuration of other measuring equipment. It can judge whether the battery system is working stably by monitoring the hydrogen production of the electrolyzer, so as to achieve timely maintenance;

[0101] The present invention is applicable to the collaborative hierarchical control system of the PEM-ALK hybrid hydrogen production DC microgrid system. On the basis of ensuring the stability of the DC microgrid bus voltage, it fully absorbs the fluctuating power generated by the wind turbine. BRIEF DESCRIPTION OF THE DRAWINGS

[0102] Figure 1 This is a structural block diagram of the PEM-ALK hybrid hydrogen production DC microgrid system of the present invention;

[0103] Figure 2 This is a structural block diagram of the collaborative hierarchical control system of the present invention;

[0104] Figure 3 This is a flow chart of power allocation by wavelet decomposition method in the present invention;

[0105] Figure 4 This is a structural diagram of the battery energy storage module and the electrolyzer hydrogen production module in the present invention;

[0106] Figure 5 The current-hydrogen production efficiency characteristic curves of the two electrolyzers in the present invention are shown;

[0107] Figure 6 This is a flow chart of the black kite algorithm in the present invention. DETAILED DESCRIPTION

[0108] The present invention is further described in detail below with reference to the accompanying drawings and specific embodiments.

[0109] like Figure 2 As shown, the present invention is a collaborative hierarchical control system suitable for a PEM-ALK hybrid hydrogen production DC microgrid system, which is used to accurately control the PEM-ALK hybrid hydrogen production DC microgrid system, including a power distribution control layer 1, a strategy management control layer 2, and an equipment control layer 3.

[0110] The DC microgrid system includes a wind turbine power generation module, a flywheel energy storage module, a battery energy storage module, an electrolyzer hydrogen production module, and an AC power grid module, which are respectively connected to the DC bus; the electrolyzer hydrogen production module includes an ALK electrolyzer hydrogen production module and a PEM electrolyzer hydrogen production module connected in parallel, and the ALK electrolyzer and the PEM electrolyzer are respectively connected to the DC bus through different Buck converters.

[0111] Specifically, if Figure 1 As shown, in the DC microgrid system, the wind turbine power generation module includes a wind turbine system and a converter, and the wind turbine system is connected to the DC bus through an AC / DC converter; the flywheel energy storage module includes a flywheel energy storage and a converter, and the flywheel energy storage is connected to the DC bus through a rectifier converter; the battery energy storage module includes a battery energy storage and a converter, and the battery energy storage is connected to the DC bus through a DC / DC converter; the AC grid module includes an AC grid and a converter, and the AC grid is connected to the DC bus through an AC / DC converter.

[0112] The power distribution control layer 1 is used to collect real-time wind speed data to obtain the wind turbine power generation power P w , in the absence of Internet communication, the wind turbine power generation P is converted into wDecomposed into a linear combination of multiple wavelet basis functions of different frequencies, reconstructed into the grid-connected power P bing , flywheel power P f and battery power P bat , under the premise of ensuring the stability of the DC bus voltage, the power balance of the DC microgrid system is achieved; at the same time, the allocated power signals are transmitted to the strategy management control layer 2 and the device control layer 3 respectively.

[0113] By collecting real-time wind speed data, the wind turbine power generation power P w The equation is

[0114]

[0115] Where,

[0116] P w is the wind turbine power generation power,

[0117] ρ is the air density,

[0118] V is the real-time wind speed,

[0119] R is the radius of the fan blade,

[0120] C p is the wind energy utilization coefficient, which is a function of the pitch angle β and the blade tip rotation ratio λ.

[0121] The wind turbine power P is converted into w (t) is decomposed into a linear combination of multiple wavelet basis functions of different frequencies and reconstructed into the grid-connected power P bing (t) and hybrid energy storage system power P hess (t), the hybrid energy storage system power P hess (t) is the flywheel power P f (t) and battery power P bat (t) and.

[0122] like Figure 3 As shown in FIG, it is a flow chart of power allocation by wavelet decomposition method.

[0123] According to the State Grid Corporation of China's standard "Technology for Wind Farm Integration into Power Systems," maximum power fluctuation rates are divided into 1-minute and 10-minute intervals, both of which must meet the limits set by the grid dispatching department. The maximum power fluctuation rate refers to the percentage of the wind farm's installed capacity relative to the peak-to-valley power fluctuations within a specific time interval.

[0124] Wavelet decomposition decomposes the original signal into a low-frequency signal A through low-pass filter and high-pass filter J and a series of high frequency signals [D1, D2, ..., DJ ], the decomposition process in wavelet decomposition is a step-by-step decomposition process from scale j+1 to scale j, and the decomposition algorithm is:

[0125]

[0126] The original signal After decomposition

[0127] and D j Reconstructible It can be expressed as:

[0128]

[0129] Where,

[0130] is the low-frequency signal of the original signal,

[0131] h k-2n is the impulse response of the low-pass filter,

[0132] D j is the high-frequency signal of the original signal,

[0133] G k-2n is the impulse response of the high-pass filter,

[0134] j is the number of decomposition levels,

[0135] n is the coefficient,

[0136] k is the coefficient.

[0137] The orthogonal and approximately symmetrical Symlets4 is used to decompose and reconstruct the signal. When the signal is decomposed to the third layer, the third layer signal can be expressed as:

[0138] P w =D1+D2+D3+A3

[0139] Where,

[0140] P w is the fan power,

[0141] D1 is a high frequency signal.

[0142] D2 is a high frequency signal,

[0143] D3 is a high frequency signal.

[0144] A3 is a low-frequency signal.

[0145] According to the wavelet decomposition characteristics, when the signal D i When the grid power fluctuation rate limit is met, signal Di ~D J 、A J That is, when signal D2 meets the grid connection requirements, the grid connection power P bing = D2+D3+A3, flywheel power P f and battery power P bat The sum is D1.

[0146] Wavelet quadratic decomposition is further used to reconstruct the D1 signal, where low-frequency power is allocated to the battery energy storage module and high-frequency power is allocated to the flywheel energy storage module. The power allocated to the battery energy storage module and the flywheel energy storage module meets the following conditions:

[0147]

[0148] Where,

[0149] P bat is the power allocated to the battery,

[0150] P bat_max is the maximum capacity of the battery,

[0151] P f is the power distributed to the flywheel,

[0152] P f_max is the maximum capacity of the flywheel.

[0153] The power distribution control layer in the present invention uses the wavelet decomposition method to ensure the stability of the DC bus voltage and convert the wind turbine power P w Decomposed into a linear combination of multiple wavelet basis functions of different frequencies, reconstructed into the grid-connected power P bing , flywheel power P f and battery power P bat , achieving power balance of DC microgrid system.

[0154] The strategy management control layer 2 is used to collect the battery power P in real time. bat , battery voltage U bat , battery current I bat , and battery SOC, and switch the battery control strategy and the working state of the electrolyzer according to the battery SOC range. The battery switches between constant voltage control mode and constant power control mode, and changes the working power of the ALK electrolyzer and PEM electrolyzer at the same time; then the working power of the ALK electrolyzer and PEM electrolyzer, the battery voltage U bat signal, and battery current I bat The signal is transmitted to the device control layer 3.

[0155] When the system is configured with loads of different capacities, even if P bat>0, the battery may be in one of three states: charging, discharging, or neither, depending on system needs or its own state. Furthermore, the battery's charge and discharge performance is significantly affected by multiple factors, including ambient temperature, humidity, and air pressure. These factors affect the internal chemical reaction rate and can induce corrosion and aging, altering its performance and service life in ways that are difficult to directly observe.

[0156] Taking the above problems into consideration, the present invention switches the battery control strategy and the working state of the electrolyzer according to the battery SOC. The specific control strategy for the battery, ALK electrolyzer and PEM electrolyzer is:

[0157] When the battery SOC is greater than 85%: If P bat >0, stop charging the battery; if P bat <0, control the battery to be in constant voltage (CVC) state; adjust the ALK electrolyzer and PEM electrolyzer at rated working power P e Run down;

[0158] When the battery SOC is greater than 80% and less than 85%, the battery is controlled to be in a constant voltage (CVC) state, and the ALK electrolyzer and PEM electrolyzer are adjusted to operate at the rated power P. e Run down;

[0159] When the battery SOC is greater than 30% and less than 80%, the battery is controlled to be in a constant power (CPC) state, and the ALK electrolyzer and PEM electrolyzer are adjusted to operate at the rated power P. e Run down;

[0160] When the battery SOC is greater than 15% and less than 30%, the battery is controlled to be in constant power (CPC) state, and the working power of the ALK electrolyzer is adjusted to 0.2P e , adjust the working power of PEM electrolyzer to 0.1P e ;

[0161] When the battery SOC is less than 15%: If P bat <0, the battery stops working; if P bat >0, the battery is controlled to be in constant power (CPC) state; both the ALK electrolyzer and the PEM electrolyzer are not working.

[0162] The specific strategies are shown in Table 1.

[0163] Table 1 Specific control strategies for batteries and electrolyzers

[0164]

[0165] The strategy management control layer in the present invention controls the battery SOC to be maintained within the specified range, and switches the battery control strategy and the working state of the electrolyzer according to the battery SOC. This not only reduces the number of charge and discharge times of the battery and significantly extends the battery service life, but also ensures that the electrolyzer and battery module can work together efficiently according to the preset control strategy.

[0166] The device control layer 3 is used to collect the DC microgrid bus voltage U in real time. dc , the power values ​​after distribution, the working power of the electrolytic cell, the battery voltage U bat , and battery current I bat According to the mathematical models of ALK electrolyzer and PEM electrolyzer, the current-hydrogen production efficiency characteristic curves of ALK electrolyzer and PEM electrolyzer are simulated, and then the only expected working voltage U corresponding to the maximum hydrogen production efficiency of ALK electrolyzer and PEM electrolyzer under different working power conditions is determined based on the current-hydrogen production efficiency characteristic curves. ref ; Then establish the fractional-order PI of Buck converter λ The control system uses fractional-order PI λ The control system is connected with the Black Kite algorithm and Buck converter, and the feedback objective function is used as the evaluation index to optimize the control parameters so that the actual working voltage U ALK , Actual working voltage of PEM electrolyzer U PEM and the expected operating voltage U ref The difference between them is minimized, achieving the goal of stabilizing the ALK electrolyzer and PEM electrolyzer at the maximum efficiency point.

[0167] Specifically, the mathematical model of the ALK electrolyzer is

[0168]

[0169] Where,

[0170] U ALK is the actual working voltage of ALK electrolyzer,

[0171] U rev is the reversible voltage,

[0172] U ohm is the resistance overvoltage caused by ohmic polarization,

[0173] U con is the concentration polarization overvoltage,

[0174] T is the working temperature of the electrolytic cell,

[0175] I is the current of the hydrogen production unit,

[0176] A eleis the cathode plate area,

[0177] r1, r2, s1, s2, s3, t1, t2, t3, α are adjustment coefficients.

[0178] Specifically, the mathematical model of the PEM electrolyzer is

[0179]

[0180] Where,

[0181] U PEM is the actual working voltage of the PEM electrolyzer,

[0182] U ocv is the open circuit voltage, which is also the minimum theoretical voltage of the PEM electrolytic cell.

[0183] U ohm is the ohmic overpotential due to the cell resistance,

[0184] U act is the overpotential due to the electrochemical reaction,

[0185] U act,a is the activation overpotential of the anode,

[0186] U act,c is the activation overpotential of the cathode.

[0187] Specifically, the hydrogen production efficiency η of both ALK electrolyzer and PEM electrolyzer can be expressed as:

[0188]

[0189] Where,

[0190] Q h is the chemical heat energy of hydrogen,

[0191] Q power is the electrical energy absorbed by the electrolytic cell,

[0192] Q heat It is the heat energy compensated to the electrolytic cell by the external heat source.

[0193] Through the above relationship, the I-η characteristic curves of ALK electrolyzer and PEM electrolyzer are obtained respectively.

[0194] like Figure 5 As shown in the I-η characteristic curve, as the input current increases, η ALK and η PEM They all show a trend of increasing first and then decreasing and have a unique η maxCurrently, the minimum operating load range of ALK electrolyzers is mostly above 20%, and PEM electrolyzers can adapt to a wide power operation range of 10% to 150%.

[0195] In the present invention, the rated operating power of the ALK electrolyzer and the PEM electrolyzer is P e When the electrolyzer needs to reduce power, both the ALK electrolyzer and the PEM electrolyzer are set to operate at the lowest power. That is, the working power of the ALK electrolyzer is 20% P e , the operating power of PEM electrolyzer is 10%P e At this time, the two electrolytic cells each correspond to a unique operating voltage that makes the electrolytic cell work most efficiently.

[0196] Specifically, the established fractional-order PI λ The transfer function expression of the control system is

[0197]

[0198] The fractional PI λ The output signal expression of the control system is

[0199] u(t)=K p e(t)+K I D -λ e(t)

[0200] Where,

[0201] G(s) is the fractional-order PI λ The transfer function of the control system,

[0202] K P is the scaling factor optimized using the Black Kite algorithm,

[0203] K I is the integral coefficient optimized using the Black Kite algorithm,

[0204] λ is the control parameter to be optimized using the Black Kite algorithm,

[0205] u(t) is fractional-order PI λ The output signal of the control system,

[0206] e(t) is the actual working voltage U of the Buck circuit at time t out and the expected operating voltage U ref Comparing the error signal obtained,

[0207] D is fractional-order PI λ The control system obtains the control signal.

[0208] In order to optimize the performance of the system, the Black Kite algorithm is introduced to optimize the Kp , K I , λ parameter. Fractional-order PI λ The control system is connected to the Black Kite algorithm module and Buck converter module respectively, and the feedback objective function is used as the indicator to p , K I , λ parameters are optimized. The specific steps are:

[0209] a) Determine the system objective function. In order to ensure that the system has both good dynamic performance and high static accuracy, the present invention adopts two evaluation indicators, which are expressed as follows:

[0210]

[0211] Where,

[0212] ITAE is an evaluation index of the system dynamic performance.

[0213] ISE is an evaluation index of the system static performance.

[0214] U ref is the desired operating voltage of the electrolyzer,

[0215] U out (t) is the actual working voltage of the electrolytic cell at time t,

[0216] T is the sampling time of the simulation.

[0217] In the present invention, the electrolytic cell voltage is U ref When ALK electrolyzer and PEM electrolyzer are both in η max state, so ALK electrolyzer and PEM electrolyzer correspond to different U ref .

[0218] The present invention takes the minimum voltage deviation and the shortest adjustment time as the two evaluation indicators of the objective function, and the objective function is

[0219]

[0220] Where,

[0221] F is the objective function. The smaller the value of F, the better the solution.

[0222] a is the weight coefficient, which is used to measure the weight of the system dynamic performance.

[0223] b is the weight coefficient, which is used to measure the weight between the static performance of the system.

[0224] ITAE is an evaluation index of the system dynamic performance.

[0225] ISE is an evaluation index of the system static performance.

[0226] U ref is the desired operating voltage of the electrolyzer,

[0227] U out (t) is the actual working voltage of the electrolytic cell at time t,

[0228] T is the sampling time of the simulation.

[0229] b) Use the Black Kite algorithm to optimize K in the system p , K I , λ parameter, the design of the black kite algorithm is inspired by the migration and predation behavior of the black kite. Figure 6 The present invention uses the flow chart of the black kite algorithm, and the specific steps are:

[0230] S1: Initialization phase, creating a set of random solutions and using a matrix to represent the position of each black kite.

[0231]

[0232] Where,

[0233] m is the number of potential solutions, that is, the number of black kites,

[0234] n is the dimension of the given problem, including K p , K I , the value of λ,

[0235] P ij is the position of the j-th dimension of the i-th black-winged kite.

[0236] Evenly distribute the position A of each black kite i , whose expression is:

[0237] A i =P min +Rand(P max -P min )

[0238] Where,

[0239] i is an integer in the interval [1, m],

[0240] P max is the upper bound of the i-th black-winged kite in the j-th dimension,

[0241] P min is the lower bound of the i-th black-winged kite in the j-th dimension,

[0242] Rand is a random number in the range [0, 1].

[0243] During the initialization process, the individual with the best fitness value is selected as the leader of the initial group, which is considered to be the optimal position A of the black-winged kite. L . At this time, the optimal solution A L The objective function F corresponding to the optimal solution best Expressed as:

[0244] A L =X(find(F best = = f(A i )))

[0245] F best =min(F(A i ))

[0246] S2: Attack behavior, including different attack behaviors for global exploration and search. The mathematical model of the black kite's attack behavior is:

[0247]

[0248]

[0249] Where,

[0250] is the position of the i-th black-winged kite in the j-th dimension in the t-th iteration,

[0251] is the position of the i-th black-winged kite in the j-th dimension in the t+1-th iteration,

[0252] r is a random number in the interval [0, 1],

[0253] p is a constant with a value of 0.9,

[0254] T is the total number of iterations,

[0255] t is the number of iterations completed so far.

[0256] Different attack behaviors are controlled by parameter p, and parameter b decreases nonlinearly with the increase of iteration number, which means that the algorithm shifts from global search to local search.

[0257] S3: Migration behavior. A hypothesis based on bird migration is proposed: if the fitness value of the current population is less than that of the random population, it indicates that it is not suitable to lead the population forward, and the leader will give up leadership and join the migrating population. Conversely, if the fitness value of the current population is greater than that of the random population, it will guide the population until it reaches its destination.

[0258] The mathematical model of the black-winged kite's migration behavior is:

[0259]

[0260] Where,

[0261] is the leader of the i-th black-winged kite in the j-th dimension in the t-th iteration, which is the current optimal solution. is the position of the i-th black-winged kite in the j-th dimension in the t-th iteration,

[0262] is the position of the i-th black-winged kite in the j-th dimension in the t+1-th iteration,

[0263] f i To represent the fitness value of any black kite in the current population,

[0264] f ri is the fitness value of the random position of the jth dimension obtained by any black kite in the tth iteration,

[0265] C(0,1) is a Cauchy mutation.

[0266] S4: Select the best individual. The individual with the best fitness value is considered the leader, which corresponds to the optimal position of the black-winged kite. The best individual selected by the black-winged kite and its corresponding objective function are:

[0267]

[0268] Where,

[0269] F best is the objective function corresponding to the optimal solution,

[0270] is the position of the i-th black-winged kite in the j-th dimension in the t-th iteration,

[0271] is the leader of the i-th black-winged kite in the j-th dimension in the t-th iteration, which is the current optimal solution.

[0272] A best The optimal position for the Black-winged Kite.

[0273] S5: Determine whether the required termination conditions and error accuracy are met. If so, the algorithm ends and outputs the optimal variables and the corresponding objective function value; otherwise, go to step S2 to continue solving.

[0274] The device control layer in the present invention establishes the fractional order PI of Buck converter λ Control system, using the Black Kite algorithm to optimize fractional-order PI λ The control parameters of the control system improve the efficiency and stability of the hybrid hydrogen production system, enhance the overall operating performance of the system, and meet the growing demand for clean energy.

[0275] The present invention divides the control of the entire DC microgrid system into three levels: power distribution control layer, strategy management control layer, and equipment control layer; the power distribution control layer and the strategy management control layer implement hierarchical control through the power signals in the system, and achieve power balance of the system while ensuring the stability of the DC bus voltage; the strategy management control layer and the equipment control layer regulate and control through the voltage and current signals of each module, effectively improving the operating efficiency and stability of the electrolyzer, while reducing the number of battery charge and discharge times, and preventing the battery from overcharging and over-discharging. This regulation mechanism based on power signals and voltage and current signals is conducive to the overall optimization and fine regulation of the system.

[0276] The above embodiments are preferred implementation modes of the present invention, but the implementation modes of the present invention are not limited to the above embodiments. Any other changes, modifications, substitutions, combinations, and simplifications that do not deviate from the spirit and principles of the present invention should be considered as equivalent replacement methods and are included in the scope of protection of the present invention.

Claims

1. A collaborative hierarchical control system suitable for a PEM-ALK hybrid hydrogen production DC microgrid system, used for precise control of the PEM-ALK hybrid hydrogen production DC microgrid system, characterized by: It includes a power allocation control layer (1), a policy management control layer (2), and a device control layer (3); The DC microgrid system includes a wind turbine power generation module, a flywheel energy storage module, a battery energy storage module, an electrolyzer hydrogen production module, and an AC power grid module, each of which is connected to a DC bus. The electrolyzer hydrogen production module includes an ALK electrolyzer hydrogen production module and a PEM electrolyzer hydrogen production module connected in parallel. The ALK electrolyzer and the PEM electrolyzer are connected to the DC bus through different Buck converters. The power distribution control layer (1) is used to collect real-time wind speed data to obtain the wind turbine power generation power P w , in the absence of Internet communication, the wind turbine power generation P is converted into w Decomposed into a linear combination of multiple wavelet basis functions of different frequencies, reconstructed into the grid-connected power P bing , flywheel power P f and battery power P bat , under the premise of ensuring the stability of the DC bus voltage, the power balance of the DC microgrid system is achieved; at the same time, the allocated power signals are transmitted to the strategy management control layer (2) and the device control layer (3) respectively; The strategy management control layer (2) is used to collect the battery power P in real time. bat , battery voltage U bat , battery current I bat , and battery SOC, switch the battery control strategy and the working state of the electrolyzer according to the battery SOC range, so that the battery switches between constant voltage control mode and constant power control mode, and at the same time changes the working power of the ALK electrolyzer and PEM electrolyzer; then change the working power of the ALK electrolyzer and PEM electrolyzer, the battery voltage U bat signal, and battery current I bat The signal is transmitted to the device control layer (3); The device control layer (3) is used to collect the DC microgrid bus voltage U in real time. dc , the power values ​​after distribution, the working power of the electrolytic cell, the battery voltage U bat , and battery current I bat According to the mathematical models of ALK electrolyzer and PEM electrolyzer, the current-hydrogen production efficiency characteristic curves of ALK electrolyzer and PEM electrolyzer are simulated, and then the only expected working voltage U corresponding to the maximum hydrogen production efficiency of ALK electrolyzer and PEM electrolyzer under different working power conditions is determined according to the current-hydrogen production efficiency characteristic curves. ref , establish the fractional-order PI of Buck converter λ The control system uses fractional-order PI λ The control system is connected with the Black Kite algorithm and Buck converter, and the feedback objective function is used as the evaluation index to optimize the control parameters so that the actual working voltage U ALK , Actual working voltage of PEM electrolyzer U PEM and the expected operating voltage U ref The difference between them is minimized, achieving the goal of stabilizing the ALK electrolyzer and PEM electrolyzer at the maximum efficiency point.

2. The collaborative hierarchical control system for a PEM-ALK hybrid hydrogen production DC microgrid system according to claim 1 is characterized in that: In the DC microgrid system, the wind turbine power generation module includes a wind turbine system and a converter, and the wind turbine system is connected to the DC bus through an AC / DC converter; the flywheel energy storage module includes a flywheel energy storage and a converter, and the flywheel energy storage is connected to the DC bus through a rectifier converter; the battery energy storage module includes a battery energy storage and a converter, and the battery energy storage is connected to the DC bus through a DC / DC converter; the AC power grid module includes an AC power grid and a converter, and the AC power grid is connected to the DC bus through an AC / DC converter.

3. The collaborative hierarchical control system for a PEM-ALK hybrid hydrogen production DC microgrid system according to claim 1 is characterized in that: In the power distribution control layer (1), the wind turbine power generation power P w The equation is Where, P w is the wind turbine power generation power, ρ is the air density, V is the real-time wind speed, R is the radius of the fan blade, C p is the wind energy utilization coefficient, which is a function of the pitch angle β and the blade tip rotation ratio λ.

4. The collaborative hierarchical control system for a PEM-ALK hybrid hydrogen production DC microgrid system according to claim 3 is characterized by: In the power distribution control layer (1), wavelet decomposition decomposes the original signal into a low-frequency signal A through low-pass filter and high-pass filter. J and a series of high frequency signals [D1, D2, ..., D J ], and the orthogonal and approximately symmetrical Symlets4 are used to decompose and reconstruct the signal. When the signal is decomposed to the third layer, the third layer signal can be expressed as: <h2 style=";text-align:left;direction:ltr">P<h2 style=";text-align:left;direction:ltr"> w <h2 style=";text-align:left;direction:ltr"> (D1+D2+D3+A3) Where, P w is the fan power, D1 is a high frequency signal. D2 is a high frequency signal. D3 is a high frequency signal, A3 is a low-frequency signal; When signal D2 meets the grid connection requirements, the grid connection power P bing = D2+D3+A3, flywheel power P f and battery power P bat The sum is D1.

5. The collaborative hierarchical control system for a PEM-ALK hybrid hydrogen production DC microgrid system according to claim 4 is characterized in that: In the power distribution control layer (1), wavelet quadratic decomposition is used to reconstruct the D1 signal, in which the low-frequency power is allocated to the battery energy storage module and the high-frequency power is allocated to the flywheel energy storage module. The power allocated to the battery energy storage module and the flywheel energy storage module meets the following conditions: Where, P bat is the power allocated to the battery, P bat_max is the maximum capacity of the battery, P f is the power distributed to the flywheel, P f_max is the maximum capacity of the flywheel.

6. The collaborative hierarchical control system for a PEM-ALK hybrid hydrogen production DC microgrid system according to claim 1 is characterized by: In the strategy management control layer (2), the specific control strategy for the battery, ALK electrolyzer, and PEM electrolyzer is: When the battery SOC is greater than 85%: If P bat >0, stop charging the battery; if P bat <0, control the battery to be in a constant voltage state; adjust the ALK electrolyzer and PEM electrolyzer at the rated working power P e Run down; When the battery SOC is greater than 80% and less than 85%, control the battery to be in a constant voltage state, adjust the ALK electrolyzer and PEM electrolyzer to the rated working power P e Run down; When the battery SOC is greater than 30% and less than 80%, the battery is controlled to be in a constant power state, and the ALK electrolyzer and PEM electrolyzer are adjusted to operate at the rated power P e Run down; When the battery SOC is greater than 15% and less than 30%, the battery is controlled to be in a constant power state and the working power of the ALK electrolyzer is adjusted to 0.2P e , adjust the working power of PEM electrolyzer to 0.1P e ; When the battery SOC is less than 15%: If P bat <0, the battery stops working; if P bat >0, the battery is controlled to be in constant power state; both the ALK electrolyzer and the PEM electrolyzer are not working.

7. The collaborative hierarchical control system for a PEM-ALK hybrid hydrogen production DC microgrid system according to claim 1 is characterized by: In the equipment control layer (3), the mathematical model of the ALK electrolyzer is Where, U ALK is the actual working voltage of ALK electrolyzer, U rev is the reversible voltage, U ohm is the resistance overvoltage caused by ohmic polarization, U con is the concentration polarization overvoltage, T is the working temperature of the electrolytic cell, I is the current of the hydrogen production unit, A ele is the cathode plate area, r1, r2, s1, s2, s3, t1, t2, t3, α are adjustment coefficients; The mathematical model of the PEM electrolyzer is Where, U PEM is the actual working voltage of the PEM electrolyzer, U ocv is the open circuit voltage, which is also the minimum theoretical voltage of the PEM electrolytic cell. U ohm is the ohmic overpotential due to the cell resistance, U act is the overpotential due to the electrochemical reaction, U act,a is the activation overpotential of the anode, U act,c is the activation overpotential of the cathode.

8. The collaborative hierarchical control system for a PEM-ALK hybrid hydrogen production DC microgrid system according to claim 7 is characterized in that: In the equipment control layer (3), the hydrogen production efficiency η of both ALK electrolyzer and PEM electrolyzer can be expressed as: Where, Q h is the chemical heat energy of hydrogen, Q power is the electrical energy absorbed by the electrolytic cell, Q heat It is the heat energy compensated to the electrolytic cell by the external heat source.

9. The collaborative hierarchical control system for a PEM-ALK hybrid hydrogen production DC microgrid system according to claim 8, characterized in that: The fractional PI λ The transfer function expression of the control system is The fractional PI λ The output signal expression of the control system is u(t)=K p e(t)+K I D -λ e(t) Where, G(s) is the fractional-order PI λ The transfer function of the control system, K P is the scaling factor optimized using the Black Kite algorithm, K I is the integral coefficient optimized using the Black Kite algorithm, λ is the control parameter to be optimized using the Black Kite algorithm, u(t) is fractional-order PI λ The output signal of the control system, e(t) is the actual working voltage U of the Buck circuit at time t out and the expected operating voltage U ref Comparing the error signal obtained, D is fractional-order PI λ The control system obtains the control signal.

10. The collaborative hierarchical control system applicable to the PEM-ALK hybrid hydrogen production DC microgrid system according to claim 9, characterized in that: In the device control layer (3), the minimum voltage deviation and the shortest adjustment time are the two evaluation indicators of the objective function. The objective function is Where, F is the objective function. The smaller the value of F, the better the solution. a is the weight coefficient, which is used to measure the weight of the system dynamic performance. b is the weight coefficient, which is used to measure the weight between the static performance of the system. ITAE is an evaluation index of the system dynamic performance. ISE is an evaluation index of the system static performance. U ref is the desired operating voltage of the electrolyzer, U out (t) is the actual working voltage of the electrolytic cell at time t, T is the sampling time of the simulation.

Images