Direct-current micro-grid energy management method based on adaptive variation Harris eagle optimization

Abstract

The invention relates to the technical field of intelligent power grid energy management, and discloses a direct-current micro-grid energy management method based on adaptive variation Harris eagle optimization, which adopts a maximum power tracking method to control the output power of PV and WECS, and adopts an AMHHO algorithm to obtain the reference power Pfcref of FC and the reference power Pbref of LB; an objective function of the AMHHO algorithm considers the hydrogen consumption cost of the FC, the aging cost of the FC and the aging cost of the LB at the same time, the SC automatically compensates the vacant load power, the reference power of the FC, the reference power of the LB and the load power of the SC are tracked and controlled by adopting the PID algorithm, and closed-loop control is formed. According to the direct-current micro-grid energy management method based on self-adaptive variation Harris eagle optimization, comprehensive economic objectives including the aging cost of a proton exchange membrane fuel cell, the aging cost of a lithium battery and the consumption cost of hydrogen are considered, the objective function with the minimum cost is constructed, and economic benefits are achieved to the maximum extent.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of smart grid energy management, and particularly relates to a direct current microgrid energy management method based on adaptive mutation Harris eagle optimization. BACKGROUND

[0002] In recent years, distributed power generation systems relying mainly on renewable energy sources such as solar and wind energy have shown a high growth trend and gradually cover the entire country. At the same time, the development momentum of hydrogen energy technology is also very strong. As a clean and efficient energy source, the demand for hydrogen fuel cells is showing a rapid growth trend.

[0003] With the rapid development of distributed power generation and renewable energy, hydrogen-electricity coupled direct current microgrid (HE-DCMG) has gradually become an important part of energy transformation and low-carbon economy. HE-DCMG combines hydrogen energy with direct current microgrid technology, not only improving the utilization rate of renewable energy, but also providing support for the flexibility and stability of the energy system.

[0004] In the prior art, HE-DCMG includes a photovoltaic system (PV), a wind energy system (WECS), a proton exchange membrane fuel cell (FC), a lithium battery (LB), a super capacitor (SC), and a load. PV and WECS, as the main renewable energy power generation units, provide power through solar and wind energy. However, due to the volatility and intermittency of renewable energy, the system needs to rely on energy storage devices to balance energy supply and demand to ensure the continuous supply of the load. In this process, FC converts hydrogen energy into electrical energy, and the byproduct is water, which has the characteristics of high efficiency and cleanliness. LB, as an energy storage device, can quickly respond to load changes and adjust the power output of the system. However, the service life of FC and LB is affected by their working conditions, especially the aging problem caused by rapid charging and discharging. How to reasonably allocate power to minimize costs and prolong battery life becomes a key problem in design.

[0005] In HE-DCMG, hydrogen consumption and device aging are two important factors affecting system economy. The hydrogen consumption of FC directly determines its operating cost, and the aging of FC and LB will gradually deteriorate with the use of time, leading to a decrease in battery efficiency and further increasing the long-term operating cost. Therefore, optimizing hydrogen consumption and aging cost becomes a core problem to improve system economy.

[0006] In HE-DCMG, existing energy management strategies based on optimization algorithms only focus on hydrogen consumption cost and state of charge (SoC) of LB and SC when considering system operation economy, ignoring the aging cost of FC and LB. This neglect to some extent leads to resource waste.

[0007] Harris hawk optimization (HHO) is a newly emerging intelligent optimization algorithm based on the hunting behavior in nature, which simulates the process of Harris hawk group hunting prey, and has strong global search ability and local development ability. However, the traditional HHO is easy to fall into local optimum when facing complex multimodal problems, and the convergence speed and solution accuracy may be limited. In order to solve these problems, the adaptive mutation Harris hawk optimization algorithm (AMHHO) introduces the ideas of differential evolution algorithm and Levy flight on the basis of HHO, and balances the search process between global and local through adaptive adjustment of search strategy and mutation operator, so as to improve the global search ability and convergence efficiency. In HE-DCMG, the adaptive mutation mechanism can dynamically adjust the search strategy of the algorithm according to the real-time running state of the system, so that the power distribution optimization process is more flexible and efficient, and adapts to the uncertainty of renewable energy output, the intermittence of energy storage equipment and the volatility of load demand. SUMMARY

[0008] The purpose of the present application is to overcome the above technical deficiencies, and to provide a DC microgrid energy management method based on adaptive mutation Harris hawk optimization, which considers the comprehensive economic objective including the aging cost of proton exchange membrane fuel cell, the aging cost of lithium battery and the hydrogen consumption cost, and constructs the objective function of cost minimization to maximize economic benefits.

[0009] To achieve the above purpose, the DC microgrid energy management method based on adaptive mutation Harris hawk optimization disclosed by the present application adopts the maximum power tracking method to control the output power of PV and WECS, and adopts the AMHHO algorithm to obtain the reference power P fc_ref and P b_ref of LB, the objective function of the AMHHO algorithm considers the hydrogen consumption cost of FC, the aging cost of FC and the aging cost of LB, the SC automatically compensates the insufficient load power, and the PID algorithm is adopted to track and control the reference power of FC, the reference power of LB and the load power of SC, forming a closed loop control.

[0010] Preferably, the method comprises the following steps:

[0011] S1) Data acquisition: real-time acquisition of the current light intensity G, output voltage V PV and output current I PV of PV, the current wind speed v w , generator speed w r , stator current i a , i b , i c of WECS, the state of charge SoC SC of SC, the state of charge SoC b of LB and the actual output voltage VFC and the actual output current I FC , the actual output voltage V b of LB b and the actual output current I SC SC , the actual voltage V dc of DC bus load and the load power P RE ;

[0012] S2) obtaining the renewable energy power P HESS_ref and the hybrid energy storage reference power P RE , the renewable energy power P PV is the sum of the output power of PV and WECS, the output power of PV P W is obtained by maximum power tracking method according to the light intensity and wind speed, and the output power of WECS P RE is obtained by maximum power tracking method according to the light intensity and wind speed, so that the renewable energy power P PV = P W + P HESS_ref , the hybrid energy storage reference power P load = P RE -P FC_ref ;

[0013] S3) energy management, using AMHHO algorithm to solve the designed objective function, realizing power distribution, obtaining the FC reference power P b_ref and the LB reference power P PV , so as to make the operation cost reach the lowest and achieve the goal of making the DC microgrid run more economically, efficiently and reliably;

[0014] S4) PID control: FC and LB track the corresponding reference power stably through the PID controller, at the same time, SC adjusts the bus voltage through the PID controller to maintain the stability of the bus voltage, and automatically compensates the difference power of the system to meet the load demand.

[0015] Preferably, the step S2) comprises the following steps:

[0016] S201) making PV reach the stable maximum power point through disturbance observation method to obtain the output power P PV of PV:

[0017] P PV = V PV · I W ;

[0018] S202) the output power P p of WECS and the power coefficient CThat is, the wind energy utilization efficiency is proportional to the optimal tip speed ratio to achieve maximum power tracking method, and through the three-phase voltage type PWM converter rectifier to power into the DC bus, power coefficient C p At a certain tip speed ratio λ opt reaches a maximum value, that is, the system through the control of wind turbine speed w r To maintain the optimal tip speed ratio, C p Maximum, thereby achieving maximum power output, tip speed ratio λ formula as follows:

[0019]

[0020] Where R is the radius of the wind turbine, v w The wind speed, adjust the generator electromagnetic torque T e Control the wind turbine speed w r For the target speed Where ψ f Is the rotor flux, i q Is the q-axis current, p is the number of pole pairs, the stator current i a , i b , i c Convert to d-q coordinate system, calculate d-axis current i d And q-axis current i q , and set i d = 0, control i q , by adjusting i q Control the electromagnetic torque T e , so as to adjust the generator output power, the d-q axis voltage equation of permanent magnet synchronous generator is:

[0021]

[0022] Where u d Is the d-axis voltage, u q Is the q-axis voltage, R s Is the stator resistance, L d Is the d-axis inductance, L q Is the q-axis inductance, w e Is the electrical angular velocity, so that WECS reaches the stable maximum power point, that is, the output power P W Of WECS:

[0023]

[0024] Where ρ is the air density, A is the wind turbine swept area, and β is the pitch angle.

[0025] S203) obtain renewable energy power P RE :

[0026] P RE= P PV + P W ;

[0027] S204) obtaining a mixed energy storage reference power P HESS_ref :

[0028] P HESS_ref = P load - P RE .

[0029] Preferably, the step S3) comprises the following steps:

[0030] S301: defining a cost objective function as equivalent hydrogen consumption cost, FC aging cost and LB aging cost, the equivalent FC hydrogen consumption being proportional to the energy level of LB and SC multiplied by a penalty factor a, the penalty factor a varying according to the state of charge of LB and SC, the penalty factor a b of LB and the penalty factor a SC of SC being as follows:

[0031]

[0032] wherein SoC b_max and SoC SC_max are the upper limits of the state of charge of LB and SC, SoC b_min and SoC SC_min are the lower limits of the state of charge of LB and SC, the equivalent hydrogen consumption cost C H2 is as follows:

[0033]

[0034] wherein n cell is the total number of batteries, m H2 is the molar mass of hydrogen, F is the Faraday constant, P FC , P b , P SC are the powers output by FC, LB and SC respectively, P rice_H2 is the price per gram of hydrogen, the degradation of FC has four forms: power change, high power, low power and start-stop, the life decay AD FC of FC is a function of the running time of each degradation condition, as follows:

[0035]

[0036] wherein d1, d2, d3, d4 are constants, T high and T idle are the working time of FC in high power and low power states respectively, N start_stope is the number of start-stop cycles of FC, P change_FCis the change of FC output power, k p is the calibration coefficient, high power and low power are defined as output power greater than 80% of rated power and output power less than 20% of rated power respectively, the aging cost of FC is as follows:

[0037]

[0038] wherein, ΔD is 0.1, P rice_FC is the price of FC, P FC_max is the maximum output power of FC, the aging cost of LB is as follows:

[0039]

[0040] wherein, P rice_b is the price of LB, P b_max is the maximum output power of LB, the target function of system operation is as follows:

[0041]

[0042] At the same time, in order to ensure the safe operation of the system, the following constraint conditions need to be met:

[0043]

[0044] wherein, P FC_min is the minimum output power of FC, P b_min is the minimum output power of LB;

[0045] S302) In order to reasonably allocate power in the system with dynamic load demand and multiple constraints, AMHHO algorithm is used to solve the target function, P FC and P b correspond to two kinds of Harris eagle population FC population and b population, P FC_i and P b_i represent the position of the i-th eagle in each population, that is, the candidate solution, and P FC_ref and P b_ref correspond to the best candidate solution of each population in iteration, which is also the position of the prey to be captured, so as to obtain the best candidate solution P FC and P b .

[0046] Preferably, in the step S302), P FC and P b The best candidate solution of iteration is obtained by the following steps, and the solution of the target function is realized:

[0047] S3021) Differential evolution algorithm is introduced:

[0048] By exponential function adjustment, the balance exploration and development ability adopts a nonlinear decay function to control the mutation probability p m and mutation range R m , so that the mutation probability p m and mutation range R m are higher in the early stage and rapidly decrease in the later stage, as shown in the following formula:

[0049]

[0050] Wherein, n represents the nth iteration, N is the maximum iteration number of the population, only when the random number r0 between 0 and 1 randomly generated by the system is less than p m , mutation operation is performed, otherwise, mutation operation is skipped, the differential evolution algorithm explores the search space by differential mutation of candidate solutions, for each individual P FC_i or P b_i , a mutated individual is generated by the following formula:

[0051] P FC_i (n+1)=P FC_r1 (n)+R m ·(P FC_r2 (n)-P FC_r3 (n))

[0052] P b_i (n+1)=P b_r1 (n)+R m ·(P b_r2 (n)-P b_r3 (n))

[0053] Wherein, P FC_r1 , P FC_r2 , P FC_r3 are the power of fuel cells randomly selected from the FC population, P b_r1 , P b_r2 , P b_r3 are the power of lithium batteries randomly selected from the b population, in each iteration, by comparing the fitness of the newly generated mutated individual and the current individual, the better individual is selected to enter the next generation;

[0054] S3022) introduce Levy flight:

[0055] The following step formula of Levy flight is used to update the power of the individual:

[0056] P FC_i (n+1)=P FC_i (n)+SεLevy

[0057] P b_i (n+1)=P b_i (n)+S·Levy

[0058]

[0059] where Levy is the step size of Levy flight, S is a d-dimensional random vector between 0 and 1, u, v are random numbers between 0 and 1, β is a constant of 1.5, σ represents the scale parameter when generating random numbers, and Γ is the gamma function;

[0060] S3023) Mutation mixed strategy:

[0061] Differential evolution mutation and Levy flight are applied simultaneously in each iteration, the obtained candidate solution is substituted into the objective function, the calculation results are compared, and then the global best candidate solution P FC_ref_1 and P b_ref_1 are updated

[0062] S3024) Local search:

[0063] Small perturbations are made around the global best candidate solution in the last 20% of iterations to obtain new candidate solutions P FC_ref_2 and P b_ref_2 :

[0064] P FC_ref_2 = P FC_ref_1 + 0.01 (P FC_max - P FC_min ) · (r1-0.5)

[0065] P b_ref_2 = P b_ref_1 + 0.01 (P b_max - P b_min ) · (r2-0.5)

[0066] where r1 and r2 are random numbers between 0 and 1;

[0067] S3025) Obtain P FC_ref and P b_ref

[0068] Finally, the candidate solutions P FC_ref_2 and P b_ref_2 are substituted into the objective function, and the calculation results are compared with the calculation results of substituting P FC_ref_1 and P b_ref_1 into the objective function, and P FC_ref and P b_ref are obtained:

[0069] If f(P FC_ref_1 , P b_ref_1 ) < f(P FC_ref_2 , P b_ref_2 ), then

[0070] P FC_ref= P FC_ref_1 , b_ref = P b_ref_1 ;

[0071] If f(P FC_ref_1 , P b_ref_1 ) ≥ f(P FC_ref_2 , P b_ref_2 ), then

[0072] P FC_ref = P FC_ref_2 , P b_ref = P b_ref_2 .

[0073] Preferably, in the step S4), the PID control comprises the following steps:

[0074] S401) obtaining the reference power P FC_ref and P b_ref of the FC and the LB from the FC and the LB in the step S3), obtaining the corresponding reference current I FC_ref and I b_ref :

[0075]

[0076] In order to make the output current I FC and I b of the FC and the LB track the corresponding reference current I FC_ref and I b_ref , the following current error variables are introduced:

[0077] e1 = I FC - I FC_ref

[0078] e2 = I b - I b_ref

[0079] The duty ratio D1 of the FC side boost converter and the duty ratio D2 of the LB side bidirectional DC converter are obtained by the PID algorithm:

[0080]

[0081] Wherein, K p_1 , K i_1 , K d_1 are the proportional, integral and differential gains of the FC side PID algorithm respectively, K p_2 , K i_2 , K d_2 are the proportional, integral and differential gains of the LB side PID algorithm respectively;

[0082] S402) In order to maintain the actual voltage V dcThe stability of the DC micro-grid is maintained at 750V, and the following voltage error variable is introduced:

[0083] e3 = V dc -750

[0084] The duty ratio D3 of the SC side bidirectional converter can be obtained by the PID controller:

[0085]

[0086] Wherein, K p_3 , K i_3 , K d_3 are the proportional, integral and differential gains of the SC side PID algorithm respectively.

[0087] A DC micro-grid, comprising a photovoltaic system PV, a wind energy system WECS, a proton exchange membrane fuel cell FC, a lithium battery LB, a super capacitor SC and a load, the photovoltaic system PV is connected to a DC bus through a boost converter, the wind energy system WECS is connected to the DC bus through an inverter, the proton exchange membrane fuel cell FC is connected to the DC bus through a boost converter, the lithium battery LB and the super capacitor SC are connected to the DC bus through bidirectional DC converters respectively, and the adaptive mutation Harris eagle optimization-based DC micro-grid energy management method is executed.

[0088] An electronic device, comprising a memory and a processor, which are in communication connection with each other, and the memory stores computer instructions, and the processor executes the computer instructions to realize the adaptive mutation Harris eagle optimization-based DC micro-grid energy management method.

[0089] A computer readable storage medium, which stores a computer program, and the computer program is executed by a processor to realize the adaptive mutation Harris eagle optimization-based DC micro-grid energy management method.

[0090] Compared with the prior art, the present application has the following advantages:

[0091] 1. Considering the aging cost of the proton exchange membrane fuel cell, the aging cost of the lithium battery and the hydrogen consumption cost, a cost minimization objective function is constructed to maximize economic benefits;

[0092] 2. Under the premise of maintaining appropriate SoC, the designed objective function not only considers the hydrogen consumption cost, but also includes the aging cost of the FC and the LB, so as to balance the service life of the FC and the LB while optimizing the economic benefits;

[0093] 3. On the basis of HHO, the differential evolution algorithm and Levy flight are introduced, the search strategy and mutation operator are adaptively adjusted, so that the AMHHO can find a power distribution scheme with lower cost and higher economic benefit, and the optimization strategy can be adjusted according to the system dynamic change and different operating conditions, so that the system can still maintain high robustness and adaptability when facing different external environments or equipment failures;

[0094] 4. The application not only considers the economy of system operation, but also considers the stability of bus voltage, in order to maintain the stability of bus voltage, the duty cycle of the converter is controlled by the PID controller to form a closed loop control, and the AMHHO is used to realize the goal of making the HE-DCMG run more economically, efficiently and reliably. BRIEF DESCRIPTION OF DRAWINGS

[0095] Figure 1 It is a structure schematic diagram of the DC microgrid in the application;

[0096] Figure 2 It is a flow chart of the AMHHO algorithm in the DC microgrid energy management method based on adaptive mutation Harris hawk optimization. DETAILED DESCRIPTION

[0097] The technical solutions of the application will be described clearly and completely in combination with the drawings. Obviously, the described embodiments are part of the embodiments of the application, rather than all the embodiments. Based on the embodiments in the application, all other embodiments obtained by those skilled in the art without creative labor belong to the protection scope of the application.

[0098] As shown in Figure 1 A DC microgrid, including a photovoltaic system PV, a wind energy system WECS, a proton exchange membrane fuel cell FC, a lithium battery LB, a super capacitor SC and a load, the photovoltaic system PV is connected to a DC bus through a boost converter, the wind energy system WECS is connected to the DC bus through an inverter, the proton exchange membrane fuel cell FC is connected to the DC bus through a boost converter, the lithium battery LB and the super capacitor SC are respectively connected to the DC bus through a bidirectional DC converter, and a DC microgrid energy management method based on adaptive mutation Harris hawk optimization is executed.

[0099] In the embodiment, the DC microgrid energy management method based on adaptive mutation Harris hawk optimization adopts a maximum power tracking method to control the output power of the PV and the WECS, and adopts an AMHHO algorithm to obtain the reference power P fc_ref of the FC and the reference power P b_refThe objective function of the AMHHO algorithm considers the hydrogen consumption cost of the FC, the aging cost of the FC and the aging cost of the LB at the same time, the SC automatically compensates for the shortage of load power, the PID algorithm is adopted to track and control the reference power of the FC, the reference power of the LB and the load power of the SC, and a closed-loop control is formed.

[0100] Specifically includes the following steps:

[0101] S1) Data acquisition: real-time acquisition of the current light intensity G of the PV, the output voltage V PV and the output current I PV , the current wind speed v w of the WECS, the generator speed w r , the stator current i a , i b , i c , the state of charge SoC of the SC SC , the state of charge SoC of the LB b , the actual output voltage V FC and the actual output current I FC of the FC, the actual output voltage V b and the actual output current I b of the LB, the actual output voltage V SC and the actual output current I SC of the SC, the actual voltage V dc and the load power P load of the DC bus;

[0102] S2) Obtain the renewable energy power P RE and the hybrid energy storage reference power P HESS_ref , the renewable energy power P RE is the sum of the output power of the PV and the WECS, the output power P PV of the PV and the output power P W of the WECS are obtained through the maximum power tracking method according to the light intensity and the wind speed, so as to obtain the renewable energy power P RE =P PV +P W , and the hybrid energy storage reference power P HESS_ref =P load -P RE ;

[0103] S3) Energy management, the AMHHO algorithm is adopted to solve the designed objective function, realize power distribution, obtain the FC reference power P FC_ref , and the LB reference power P b_ref ;

[0104] S4) PID control: FC and LB track the corresponding reference power stably through the PID controller, at the same time, SC adjusts the bus voltage through the PID controller to maintain the stability of the bus voltage, and automatically compensates the difference power of the system to meet the load demand.

[0105] Wherein, step S2) comprises the following steps:

[0106] S201) Through the disturbance observation method, the PV reaches the stable maximum power point, and the output power P PV of the PV is obtained

[0107] P PV = V PV ·I PV ;

[0108] S202) The output power P W of the WECS is proportional to the power coefficient C p , that is, the wind energy utilization efficiency, the maximum power tracking method is adopted by adopting the best tip speed ratio, and the power is merged into the DC bus through the three-phase voltage type PWM converter rectifier, the power coefficient C p reaches the maximum value at a certain tip speed ratio λ opt , that is, the system controls the wind wheel speed w r to maintain the best tip speed ratio, so that C p is maximum, and then the maximum power output is realized, and the tip speed ratio λ formula is as follows:

[0109]

[0110] Wherein, R is the radius of the wind wheel, v w is the wind speed, the generator electromagnetic torque T e is adjusted, the wind wheel speed w r is controlled as the target speed Wherein, ψ f is the rotor flux, i q is the q-axis current, and p is the pole pair number. The stator current i a , i b and i c are converted into d-q coordinate system, the d-axis current i d and the q-axis current i q are calculated, i d =0 is set, i q is controlled, the electromagnetic torque T q is controlled by adjusting i e , so as to adjust the generator output power, and the d-q axis voltage equation of the permanent magnet synchronous generator is:

[0111]

[0112] Wherein, ud is the d-axis voltage, u q is the q-axis voltage, R s is the stator resistance, L d is the d-axis inductance, L q is the q-axis inductance, w e is the electrical angular velocity, w W :

[0113]

[0114] where p is the air density, A is the swept area of the wind turbine, and b is the pitch angle;

[0115] S203) obtaining the renewable energy power P RE :

[0116] P RE = P PV + P W ;

[0117] S204) obtaining the hybrid energy storage reference power P HESS_ref :

[0118] P HESS_ref = P load - P RE .

[0119] Step S3) comprises the following steps:

[0120] S301: defining a cost objective function as an equivalent hydrogen consumption cost, an FC aging cost, and an LB aging cost, the equivalent FC hydrogen consumption being proportional to the energy levels of the LB and the SC multiplied by a penalty factor a, the penalty factor a varying according to the state of charge of the LB and the SC, the penalty factor a b of the LB and the penalty factor a SC of the SC being as follows:

[0121]

[0122] where SoC b_max and SoC SC_max are the upper limits of the state of charge of the LB and the SC, SoC b_min and SoC SC_min are the lower limits of the state of charge of the LB and the SC, the equivalent hydrogen consumption cost C H2 is as follows:

[0123]

[0124] where n cell is the total number of batteries, and m H2is the molar mass of hydrogen, F is Faraday's constant, P FC , P b , P SC are the power output of FC, LB and SC respectively, P rice_H2 is the price per gram of hydrogen, FC has four forms of degradation: power variation, high power, low power and start-stop, the life decay of FC is ΔD FC is a function of the running time of each degraded condition, as follows:

[0125]

[0126] where d1, d2, d3, d4 are constants, T high and T idle are the working time of FC in high power and low power state respectively, N start_stope is the number of start-stop cycles of FC, P change_FC is the variation of FC output power, k p is the calibration coefficient, high power and low power are defined as output power greater than 80% of rated power and output power less than 20% of rated power respectively, the aging cost of FC is as follows:

[0127]

[0128] where ΔD is 0.1, P rice_FC is the price of FC, P FC_max is the maximum output power of FC, the aging cost of LB is as follows:

[0129]

[0130] where P rice_b is the price of LB, P b_max is the maximum output power of LB, the objective function of system operation is as follows:

[0131]

[0132] At the same time, in order to ensure the safe operation of the system, the following constraints need to be met:

[0133]

[0134] where P FC_min is the minimum output power of FC, P b_min is the minimum output power of LB;

[0135] S302) In order to reasonably allocate power in the system with dynamic load demand and multiple constraints, AMHHO algorithm is used to solve the objective function, combined with Figure 2 , P FC and Pb P FC_i and P b_i represent the position of the i-th hawk in the respective population, i.e. the candidate solution, while P FC_ref and P b_ref correspond to the best candidate solution of the respective population at iteration, i.e. the position of the prey to be captured, so as to find the best candidate solution P FC and P b .

[0136] In particular, in step S302) P FC and P b are found by the following steps, which allow to solve the objective function:

[0137] S3021 ) introducing the differential evolution algorithm:

[0138] By exponential function adjustment, the exploration and exploitation ability is balanced, and a nonlinear decay function is used to control the mutation probability p m and the mutation range R m , so that the mutation probability p m and the mutation range R m are higher in the early stage and decrease rapidly in the later stage, as shown in the following formula:

[0139]

[0140] where n represents the n-th iteration, N is the maximum number of iterations of the population, and only when the random number r0between 0 and 1 generated by the system is less than p m , the mutation operation is performed, otherwise the mutation operation is skipped. The differential evolution algorithm explores the search space by differential mutation of the candidate solution. For each individual P FC_i or P b_i , a mutated individual is generated by the following formula:

[0141] P FC_i (n+1) = P FC_r1 (n) + R m ·(P FC_r2 (n) - P FC_r3 (n))

[0142] P b_i (n+1) = P b_r1 (n) + R m ·(P b_r2 (n) - P b_r3 (n))

[0143] where P FC_r1 , P FC_r2 , P FC_r3P is the randomly selected fuel cell power in the FC population, P b_r1 b_r2 b_r3 P is the randomly selected lithium battery power in the b population, in each iteration, the better individual is selected into the next generation by comparing the fitness of the newly generated mutant individual and the current individual;

[0144] S3022) Introduce Levy flight:

[0145] The power of the individual is updated using the following step formula of Levy flight:

[0146] P FC_i (n+1) = P FC_i (n) + S Levy

[0147] P b_i (n+1) = P b_i (n) + S Levy

[0148]

[0149] Where Levy is the step length of Levy flight, S is a d-dimensional random vector between 0 and 1, u, v are random numbers between 0 and 1, β is a constant of 1.5, σ represents the scale parameter when generating random numbers, and Γ is the gamma function;

[0150] S3023) Mutation hybrid strategy:

[0151] In each iteration, differential evolution mutation and Levy flight are applied at the same time, the obtained candidate solution is substituted into the objective function, the calculation results are compared, and then the global best candidate solution P FC_ref_1 and P b_ref_1 are updated;

[0152] S3024) Local search:

[0153] In the last 20% of iterations, small perturbations are made around the global best candidate solution to obtain new candidate solutions P FC_ref_2 and P b_ref_2 :

[0154] P FC_ref_2 = P FC_ref_1 + 0.01 (P FC_max - P FC_min ) · (r1-0.5)

[0155] P b_ref_2 = P b_ref_1 + 0.01 (P b_max - P b_min ) · (r2-0.5)

[0156] ​​Among them, r1 and r2 are random numbers between 0 and 1;

[0157] S3025) Obtain P FC_ref and P b_ref

[0158] Finally, the candidate solution P FC_ref_2 and P b_ref_2 Substitute into the objective function, and P FC_ref_1 and P b_ref_1 Substitute the calculated results of the objective function for comparison and obtain P FC_ref and P b_ref :

[0159] If f(P FC_ref_1 、P b_ref_1 ) <f(P FC_ref_2 、P b_ref_2 ),but

[0160] P FC_ref =P FC_ref_1 ,P b_ref =P b_ref_1 ;

[0161] If f(P FC_ref_1 、P b_ref_1 )≥f(P FC_ref_2 、P b_ref_2 ),but

[0162] P FC_ref =P FC_ref_2 ,P b_ref =P b_ref_2 .

[0163] Finally, in step S4), PID control includes the following steps:

[0164] S401) The reference power P of FC and LB is obtained from step S3). FC_ref and P b_ref , and obtain the corresponding reference current I FC_ref and I b_ref :

[0165]

[0166] In order to make the output current I FC and I b Track the corresponding reference current I FC_ref and I b_ref , introduce the following current error variable:

[0167] e1=I FC -I FC_ref

[0168] e2=Ib -I b_ref

[0169] The duty ratio D1 of the FC side boost converter and the duty ratio D2 of the LB side bidirectional DC converter are obtained by the PID algorithm:

[0170]

[0171]

[0172] Wherein, K p_1 , K i_1 , K d_1 are the proportional, integral and differential gains of the FC side PID algorithm respectively, K p_2 , K i_2 , K d_2 are the proportional, integral and differential gains of the LB side PID algorithm respectively.

[0173] S402) In order to maintain the actual voltage V dc of the DC bus at 750V, the following voltage error variable is introduced:

[0174] e3 = V dc - 750

[0175] The duty ratio D3 of the SC side bidirectional converter can be obtained by the PID controller:

[0176]

[0177] Wherein, K p_3 , K i_3 , K d_3 are the proportional, integral and differential gains of the SC side PID algorithm respectively.

[0178] In addition, the embodiment also includes an electronic device, including a memory and a processor, the memory and the processor are mutually communicated and connected, the memory stores computer instructions, and the processor realizes the DC microgrid energy management method based on adaptive mutation Harris eagle optimization by executing the computer instructions. A computer readable storage medium, the computer readable storage medium stores a computer program, and the computer program is executed by the processor to realize the DC microgrid energy management method based on adaptive mutation Harris eagle optimization.

[0179] The application is based on a direct-current microgrid energy management method of adaptive mutation Harris hawk optimization, considers the comprehensive economic target of proton exchange membrane fuel cell aging cost, lithium battery aging cost and hydrogen consumption cost, constructs a cost minimization objective function, and maximizes economic benefits; under the premise of maintaining proper SoC, the designed objective function not only considers the hydrogen consumption cost, but also includes the aging cost of FC and LB, so as to balance the service life of FC and LB while optimizing economic benefits; on the basis of HHO, differential evolution algorithm and Levy flight are introduced, the search strategy and mutation operator are adaptively adjusted, AMHHO can find a power distribution scheme with low cost and high economic benefit, and the optimization strategy can be adjusted according to the dynamic change of the system and different operating conditions, so that the system can maintain high robustness and adaptability when facing different external environments or equipment failures; the application not only considers the economy of the system operation, but also considers the stability of the bus voltage, uses a PID controller to control the duty cycle of the converter to form a closed-loop control, and realizes the goal of making HE-DCMG run more economically, efficiently and reliably together with AMHHO.

[0180] Meanwhile, it should be noted that the description of the above technical solutions is exemplary, and the present specification can be embodied in different forms, and should not be interpreted as being limited to the technical solutions set forth herein. On the contrary, providing these descriptions will make the present disclosure thorough and complete, and will fully convey the scope disclosed by the present specification to those skilled in the art. In addition, the technical solutions of the present application are limited by the scope of the claims. The features of various embodiments of the present application can be partially or entirely combined or spliced with each other, and can be executed in various different ways as those skilled in the art can fully understand. The embodiments of the present application can be executed independently of each other, or can be executed together in a mutually dependent relationship.

[0181] For those skilled in the art to which the present application belongs, without departing from the concept of the present application, a number of simple deductions or replacements can be made, which should be considered as falling within the scope of protection of the present application.

Claims

1. A DC microgrid energy management method based on adaptive mutation Harris hawk optimization, characterized in that: The maximum power tracking method is used to control the output power of the PV and the WECS, and the AMHHO algorithm is used to obtain the reference power P fc_ref of the FC b_ref The objective function of the AMHHO algorithm considers the hydrogen consumption cost of the FC, the aging cost of the FC and the aging cost of the LB, the SC automatically compensates for the shortage of the load power, the PID algorithm is used to track and control the reference power of the FC, the reference power of the LB and the load power of the SC, and a closed-loop control is formed.

2. The DC microgrid energy management method based on adaptive mutation Harris hawk optimization of claim 1, wherein: Specifically comprising the following steps: S1) Data collection: real-time collection of PV's current light intensity G, output voltage V PV and output current I PV , the current wind speed v of WECS w , generator speed w r , stator current i a 、i b 、i c , SC's state of charge SoC SC , LB's state of charge SoC b , the actual output voltage of FC is V FC and the actual output current I FC , the actual output voltage of LB is V b and the actual output current I b , the actual output voltage of SC is V SC and the actual output current I SC , the actual voltage of the DC bus V dc and load power P load ; S2) obtaining a renewable energy power P RE and a hybrid energy storage reference power P HESS_ref , the renewable energy power P RE is the sum of the output power of PV and WECS, the output power of PV P PV and WECS P W is obtained by maximum power tracking method according to the light intensity and wind speed respectively, so as to obtain the renewable energy power P RE = P PV + P W , the hybrid energy storage reference power P HESS_ref = P load - P RE ; S3) Energy management, the AMHHO algorithm is used to solve the designed objective function, power distribution is realized, and the FC reference power P FC_ref is obtained b_ref ; S4) PID control: FC and LB are stably tracked by the PID controller corresponding to the reference power, at the same time, SC adjusts the bus voltage by the PID controller to maintain the stability of the bus voltage, and automatically compensates the difference power of the system to meet the load demand.

3. The DC microgrid energy management method based on adaptive mutation Harris hawk optimization of claim 2, wherein: The step S2) comprises the following steps: S201) By the perturbation observation method, make the PV reach the stable maximum power point, obtain the output power P of the PV PV : P PV = V PV I PV ; S202) WECS's output power P W is proportional to the wind energy utilization efficiency, the maximum power tracking method is realized by adopting the optimal tip speed ratio, and the power is merged into the DC bus through the three-phase voltage type PWM converter rectification. The power coefficient C p p At a certain tip speed ratio λ opt , reaches the maximum value, that is, the system maintains the optimal tip speed ratio by controlling the wind wheel speed w r to make C p maximum, thereby realizing the maximum power output. The tip speed ratio λ formula is as follows:​ where R is the wind wheel radius, v w is the wind speed, the electromagnetic torque T e of the generator is adjusted r The target speed w is controlled f where ψ q is the rotor flux, i a is the q-axis current, p is the number of pole pairs, the stator current i b is converted to the d-q coordinate system, the d-axis current i c and the q-axis current i d are calculated q , and i d =0 is set, i q is controlled, the electromagnetic torque T q is controlled by adjusting i e , so as to adjust the output power of the generator, and the d-q axis voltage equation of the permanent magnet synchronous generator is: where u d is the d-axis voltage, u q is the q-axis voltage, R s is the stator resistance, L d is the d-axis inductance, L q is the q-axis inductance, w e is the electrical angular speed, such that the WECS reaches the maximum power point at which the output power P W of the WECS is obtained. Wherein, ρ is the air density, A is the wind wheel swept area, and β is the pitch angle; S203) obtaining a renewable energy power P RE : P RE = P PV + P W ; S204) obtaining a mixed energy storage reference power P HESS_ref : P HESS_ref = P load - P RE .

4. The DC microgrid energy management method based on adaptive mutation Harris hawk optimization of claim 3, wherein: The step S3) comprises the following steps: S301: define the cost objective function as equivalent hydrogen consumption cost, FC aging cost and LB aging cost, equivalent FC hydrogen consumption is proportional to the energy level of LB and SC multiplied by a penalty factor a, the penalty factor a varies according to the state of charge of LB and SC, the penalty factor a of LB b and the penalty factor a of SC SC as follows: where SoC b_max and SoC SC_max are the upper limits of the LB and SC state of charge, respectively, SoC b_min and SoC SC_min are the lower limits of the LB and SC state of charge, respectively, and the equivalent hydrogen consumption cost C H2 is given by: where n cell is the total number of batteries, m H2 is the molar mass of hydrogen, F is the Faraday constant, P FC , P b , P SC are the power outputs of FC, LB and SC, respectively, P rice_H2 is the price per gram of hydrogen, FC degradation has four forms: power variation, high power, low power and start-stop, the life decay of FC ΔD FC is a function of the running time of each degradation condition, as follows: where d1, d2, d3, d4 are constants, T high and T idle are the operating time of the FC in high and low power states, respectively, N start_stope is the number of start-stop cycles of the FC, P change_FC is the variation of the FC output power, k p is the calibration factor, high and low power are defined as output power greater than 80% of the rated power and output power less than 20% of the rated power, respectively, and the aging cost of the FC is as follows: where ΔD is 0.1, P rice_FC is the price of FC, P FC_max is the maximum output power of FC, and the aging cost of LB is as follows: Among them, P rice_b is the price of LB, P b_max is the maximum output power of LB, and the objective function of the system operation is as follows: At the same time, in order to ensure the safe operation of the system, the following constraint conditions need to be met: where P FC_min is the minimum output power of the FC, P b_min is the minimum output power of the LB; S302) To allocate power reasonably in the system with dynamic load demand and multiple constraints, the AMHHO algorithm is used to solve the objective function, P FC and P b correspond to two kinds of Harris eagle populations FC population and b population respectively, P FC_i and P b_i represent the position of the i-th eagle in the respective population, that is, the candidate solution, and P FC_ref and P b_ref correspond to the best candidate solution of each population in iteration, which is also the position of the prey to be captured, so as to obtain the best candidate solution P FC and P b .

5. The DC microgrid energy management method based on adaptive mutation Harris hawk optimization of claim 4, wherein: In said step S302), P FC and P b The solution of the objective function is obtained by finding the best candidate solution of the iteration by the following steps: S3021) introducing differential evolution algorithm: By exponential function adjustment, the balance of exploration and exploitation ability adopts a nonlinear decay function to control the mutation probability p m and mutation range R m , so that the mutation probability p m and mutation range R m are high in the early stage and rapidly decrease in the later stage, as shown in the following formula: where n represents the nth iteration, N is the maximum iteration number of the population, only when the random number r0 between 0 and 1 randomly generated by the system is less than p m , the mutation operation is performed, otherwise, the mutation operation is skipped, the differential evolution algorithm explores the search space by differential mutation of the candidate solution, for each individual P FC_i or P b_i , a mutated individual is generated by the following formula: P FC_i (n+1) = P FC_r1 (n) + R m · (P FC_r2 (n) - P FC_r3 (n)) P b_i (n+1) = P b_r1 (n) + R m · (P b_r2 (n) - P b_r3 (n)) where P FC_r1 , P FC_r2 , P FC_r3 is a randomly selected fuel cell power from the FC population, P b_r1 , P b_r2 , P b_r3 is a randomly selected lithium battery power from the b population, and in each iteration, the better individual is selected into the next generation by comparing the fitness of the newly generated mutated individual and the current individual. S3022) introducing Levy flight: The following step length formula of Levy flight is used to update the power of the individual: P FC_i (n+1) = P FC_i (n) + S Levy P b_i (n+1) = P b_i (n) + S Levy Wherein, Levy is the step length of Levy flight, S is a d-dimensional random vector between 0 and 1, u and v are random numbers between 0 and 1, β is a constant of 1.5, σ represents the scale parameter when generating random numbers, and Γ is the gamma function; S3023) mutation mixed strategy: In each iteration, the differential evolution mutation and Levy flight are applied simultaneously, the resulting candidate solutions are substituted into the objective function, the calculation results are compared, and then the global best candidate solution P is updated FC_ref_1 and P b_ref_1 ; S3024) local search: In the last 20% of iterations, small perturbations are made around the global best candidate solution to obtain a new candidate solution P FC_ref_2 and P b_ref_2 : P FC_ref_2 = P FC_ref_1 + 0.01(P FC_max - P FC_min ) · (r1- 0.5) P b_ref_2 = P b_ref_1 + 0.01(P b_max - P b_min )·(r2-0.5) Wherein, r1 and r2 are random numbers between 0 and 1; S3025) obtain P FC_ref and P b_ref Finally, the candidate solution P FC_ref_2 and P b_ref_2 into the objective function, and the calculation results of P FC_ref_1 and P b_ref_1 into the objective function are compared to obtain P FC_ref and P b_ref : If f(P FC_ref_1 , P b_ref_1 ) < f(P FC_ref_2 , P b_ref_2 ), then P FC_ref = P FC_ref_1 , P b_ref = P b_ref_1 ; If f(P FC_ref_1 , P b_ref_1 ) ≥ f(P FC_ref_2 , P b_ref_2 ), then P FC_ref = P FC_ref_2 , P b_ref = P b_ref_2 .

6. The DC microgrid energy management method based on adaptive mutation Harris hawk optimization of claim 4, wherein: In the step S4), the PID control comprises the following steps: S401) obtaining the reference power P from the FC and LB obtained in step S3) FC_ref and P b_ref , obtaining the corresponding reference current I FC_ref and I b_ref : In order to make the output current I FC and I b corresponding to the reference current I FC_ref and I b_ref , the following current error variable is introduced: e1 = I FC - I FC_ref e2 = I b - I b_ref The duty cycle D1 of the FC side boost converter and the duty cycle D2 of the LB side bidirectional DC converter are obtained by the PID algorithm: wherein K p_1 , K i_1 , K d_1 are the proportional, integral and derivative gains of the FC-side PID algorithm, respectively, K p_2 , K i_2 , K d_2 are the proportional, integral and derivative gains of the LB-side PID algorithm, respectively. S402) In order to maintain the actual voltage V dc of the DC link at 750 V, the following voltage error variable is introduced: e3 = V dc -750 The duty cycle D3 of the SC side bidirectional converter can be obtained by the PID controller: Among them, K p_3 , K i_3 , K d_3 They are the proportional, integral and differential gains of the PID algorithm on the SC side.

7. A DC microgrid, characterized by: It comprises a photovoltaic system PV, a wind energy system WECS, a proton exchange membrane fuel cell FC, a lithium battery LB, a super capacitor SC and a load, the photovoltaic system PV is connected to the DC bus through a boost converter, the wind energy system WECS is connected to the DC bus through an inverter, the proton exchange membrane fuel cell FC is connected to the DC bus through a boost converter, the lithium battery LB and the super capacitor SC are connected to the DC bus through bidirectional DC converters respectively, and the DC microgrid energy management method based on adaptive mutation Harris eagle optimization as claimed in any one of claims 1-6 is executed.

8. An electronic device, comprising: It comprises a memory and a processor, which are communicatively connected with each other, the memory stores computer instructions, and the processor executes the computer instructions to realize the DC microgrid energy management method based on adaptive mutation Harris eagle optimization as claimed in any one of claims 1-6.

9. A computer-readable storage medium, the computer-readable storage medium storing a computer program, characterized in that: The computer program is executed by the processor to realize the DC microgrid energy management method based on adaptive mutation Harris eagle optimization as claimed in any one of claims 1-6.

Images