An Energy Management Method for a Hydrogen-Electric Hybrid Power System Oriented to Anti-Aging

Abstract

The present invention discloses an energy management method for a hydrogen-electric hybrid power system oriented to anti-aging. It includes: 1) Collecting real-time information of the driving condition and generating predicted information of the driving condition within the prediction horizon; 2) Constructing a fuel cell aging model and a vehicle dynamics model based on the dynamic characteristics of the hydrogen-electric hybrid power system, and establishing an objective function at time k; 3) According to the real-time information of the driving condition, the predicted information of the driving condition, and the constraint equations, using the sequential quadratic programming algorithm to solve the objective function to obtain a current control signal, so as to control the driving condition of the vehicle in the current prediction horizon; 4) Collecting the real-time information of the driving condition at time k+N in the current prediction horizon and generating predicted information of the driving condition in the next prediction horizon; repeating steps 2) and 3) to realize the rolling update of the current control signal. The present invention realizes the safety guarantee of the hybrid power system and the load tracking requirement, which is beneficial to reducing losses and prolonging the service life.

Description

Technical Field

[0001] The present invention belongs to an energy management method in the field of vehicle intelligent control, and particularly relates to an energy management method for an in-vehicle hydrogen-electric hybrid system oriented to anti-aging. Background Art

[0002] In the context of climate change and sustainable development becoming international mainstream issues, new energy technologies represented by fuel cell systems are being applied more and more widely. In particular, fuel cell vehicles have the advantages of being clean, environmentally friendly, energy-saving and efficient, and are one of the most promising development directions in the field of new energy vehicles. However, the durability bottleneck of the current fuel cell stack is still one of the key obstacles for the large-scale commercial promotion and application of fuel cell vehicles. Therefore, the long-term durability research on the catalyst layer of non-high operating temperature fuel cells such as proton exchange membrane fuel cells has important scientific significance.

[0003] The carbon-supported platinum catalyst of proton exchange membrane fuel cells is usually used to increase the rate of oxidation-reduction reactions. If the platinum particles are severely corroded during the service life of the fuel cell, it will lead to the loss of electrochemical activity, and then directly reduce the fuel cell efficiency and service life.

[0004] In order to extend the service life of the fuel cell as much as possible and meet the dynamic characteristics and system limitations, it is of great significance to fully understand the degradation process of the catalyst platinum, effectively utilize the advantages of the hybrid power system, manage the power output distribution between the main energy source and the auxiliary energy source, and improve the durability and stability of the vehicle. Summary of the Invention

[0005] The purpose of the present invention is to provide an energy management method for a hydrogen-electric hybrid system oriented to anti-aging in view of the deficiencies of the prior art.

[0006] The solution adopted by the present invention is as follows:

[0007] The present invention includes the following steps:

[0008] 1) Collect the real-time driving condition information of the vehicle at the current moment k in real time, and use the partial least squares method to perform state prediction based on the real-time driving condition information to generate the predicted driving condition information within the prediction time domain [k + 1, k + N].

[0009] 2) Construct a fuel cell aging model and a vehicle dynamics model based on the dynamic characteristics of the hydrogen-electric hybrid system, and establish an objective function and corresponding constraint equations at the k moment according to the fuel cell aging model and the vehicle dynamics model.

[0010] 3) According to the real-time information of the driving condition, the predicted information of the driving condition, and the constraint equation, use the sequential quadratic programming algorithm to solve the objective function, obtain the output control current signal of the DC converter in the hydrogen-electric hybrid system, and drive the vehicle to travel in the current prediction horizon according to the output control current signal of the DC converter;

[0011] 4) Collect the real-time information of the vehicle's driving condition at the k+N moment in the current prediction horizon and generate the predicted information of the driving condition in the next prediction horizon; repeat steps 2) and 3) to realize the rolling update of the output current control signal of the DC converter.

[0012] The real-time information of the driving condition in step 1) includes the voltage and current of the fuel cell at the k moment, the output voltage and current of the DC converter, the voltage, current, and state of charge of the lithium battery, and the real-time load current.

[0013] In step 1), based on the real-time information of the driving condition, use the partial least squares method for state prediction to generate the predicted information of the driving condition in the prediction horizon [k+1, k+N]. The calculation formula is as follows. The predicted information of the driving condition is composed of the predicted load current in the prediction horizon [k+1, k+N]:

[0014]

[0015] Among them, represents the predicted load current at the k+1 moment, represents the predicted load current at the k+1 moment, N represents the prediction step size, I L (k) represents the real-time load current at the k moment, L1 represents the first fitting coefficient, L2 represents the second fitting coefficient, I L (k-m1) represents the real-time load current at the k moment at the previous m1 moment, and T represents the matrix transpose operation.

[0016] The formula of the objective function at the k moment in step 2) is as follows:

[0017]

[0018] Among them, J(k) represents the total battery aging loss value, N represents the prediction step size, χ(i) represents the system optimization variable at the i moment, H and f are the quadratic term coefficient matrix and the linear term coefficient matrix respectively, T represents the matrix transpose operation, and i represents the time sequence number; represents the system optimization variable when the total battery aging loss value takes the minimum value;

[0019] The formula of the constraint equation corresponding to the objective function at the k moment is as follows:

[0020] χ min ≤χ(i)≤χmax

[0021] χ(i) = [X(i), u(i)] T

[0022] X(i) = [I DC (i), V s (i), V f (i), V soc (i), r(i), c(i), θ(i)] T

[0023] X nl (i) = [r(i), c(i), θ(i)] T

[0024] X l (i) = [I DC (i), V s (i), V f (i), V soc (i)] T

[0025] u(i) = ΔI DC (i)

[0026] X l (i + 1) = AX l (i) + B u u(i) + B d d(i)

[0027] X nl (i + 1) = g(X nl (i), u(i))

[0028] Where χ min and χ max respectively represent the upper and lower bound matrices of the system state variables, X(i) represents the system state variable at the i-th moment, u(i) represents the system control variable at the i-th moment, I DC (i) is the output current of the DC / DC at the i-th moment, X nl (i) represents the microscopic state variable related to aging at the i-th moment, V s (i), V f (i) respectively represent the voltages of two capacitors in the lithium battery equivalent circuit model at the i-th moment, V soc (i) represents the percentage unit voltage mapping of the state of charge of the lithium battery at the i-th moment, X l (i) represents the system circuit state variable at the i-th moment, r(i) represents the average radius of platinum particles at the i-th moment, c(i) represents the platinum ion concentration at the i-th moment, θ(i) represents the site ratio at the i-th moment; ΔIDC (i) represents the difference between the output current of the DC converter at the next moment of the i-th moment and the output current of the DC converter at the i-th moment, A, B u , B d are the linear coefficient matrix, the control variable coefficient matrix, and the perturbation coefficient matrix respectively; X l (i + 1) represents the system circuit state variable at the (i + 1)-th moment, d(i) represents the predicted load current at the i-th moment, X nl (i + 1) represents the microscopic state variable related to aging at the (i + 1)-th moment, g(X nl (i), u(i)) represents the non-linear function of the aging kinetics model of platinum at the i-th moment with respect to the microscopic state variable X nl (i) and the system control variable.

[0029] The quadratic term coefficient matrix H, the linear term coefficient matrix f, the linear coefficient matrix A, the control variable coefficient matrix B u and the perturbation coefficient matrix B d The calculation formulas are as follows:

[0030]

[0031]

[0032]

[0033] B u = [T0 0 0 0] T

[0034]

[0035] where, diag( ) represents the diagonal matrix, 0 1×3 represents the zero vector with a dimension of 1×3, 0 1×5 represents the zero vector with a dimension of 1×5, SoC m represents the stable reference value of the state of charge, δ m is the maximum deviation, α and β respectively represent the first and second weight coefficients, r(k - 1) represents the average radius of platinum particles at the (k - 1)-th moment, 0 1×6 represents the zero vector with a dimension of 1×6, C s , C f and C b respectively represent the capacitance values of the slow dynamic capacitor, the fast dynamic capacitor, and the lithium-ion capacitor, T0 is the sampling time, R s , R f and R sd respectively represent the slow dynamic impedance, the fast dynamic impedance, and the internal impedance.

[0036] The beneficial effects of the present invention are as follows:

[0037] Based on the model, the present invention performs real-time and effective management of the power output of the energy sources in the system, ensuring that the dynamic characteristics of the power output of the vehicle fuel cell system can delay the platinum degradation under high-fluctuation actual working conditions to a certain extent, thereby achieving an anti-aging effect. And it ensures that the state of charge of the auxiliary energy lithium battery is within a certain safe range, preventing overcharging and over-discharging phenomena, thereby realizing dynamic tracking of the working condition requirements and ensuring the stability of the system. By adjusting the prediction step size, the amount of calculation that can be completed for real-time calculation and online estimation can be achieved; the power output of each energy source in the system can be effectively managed under different working conditions, thereby meeting the dynamic characteristic limitations of the system and ensuring its stability. Brief Description of the Drawings

[0038] Figure 1 It is a schematic diagram of the topological structure of the vehicle hybrid power system in the present invention.

[0039] Figure 2 It is a schematic diagram of the structure of the energy management control strategy in the present invention.

[0040] Figure 3 It is a graph of the changes in load current, output current, and lithium battery current, and a graph of the change in state of charge in the embodiment of the present invention.

[0041] Figure 4 It is a graph of the change in electrochemically active area in the embodiment of the present invention. Detailed Embodiment

[0042] The present invention will be described in detail below with reference to the drawings and specific embodiments.

[0043] As Figure 1 shown, the hydrogen-electric hybrid system involved in the present invention mainly consists of a fuel cell system, a lithium battery system, a DC converter, and an electric motor. Among them, the fuel cell system responds passively through the output current command of the DC converter, and finally the output currents of the DC converter and the lithium battery are aggregated together to supply power to the load motor to enable the vehicle to drive normally.

[0044] As Figure 2 shown, the present invention includes the following steps:

[0045] 1) Real-time collect the real-time information of the driving condition of the vehicle at the current moment k, and use the partial least squares method to perform state prediction based on the real-time information of the driving condition to generate the predicted information of the driving condition within the prediction time domain [k + 1, k + N];

[0046] The real-time driving condition information in step 1) includes the voltage and current of the fuel cell at time k, the output voltage and current of the DC converter, the voltage, current and state of charge of the lithium battery, and the real-time load current.

[0047] In step 1), based on the real-time driving condition information, the partial least squares method is used for state prediction to generate the predicted driving condition information within the prediction time domain [k + 1, k + N]. The calculation formula is as follows, and the predicted driving condition information is composed of the predicted load current within the prediction time domain [k + 1, k + N]:

[0048]

[0049] Among them, represents the predicted load current at time k + 1, represents the predicted load current at time k + 1, N represents the prediction step, I L (k) represents the real-time load current at time k, L1 represents the first fitting coefficient, L2 represents the second fitting coefficient, and the first and second fitting coefficients are obtained by fitting based on historical load current data, I L (k - m1) represents the real-time load current at time k - m1 before time k, and represents the memory length. In specific implementation, the memory length is taken as 10. T represents the matrix transpose operation.

[0050] 2) Based on the dynamic characteristics of the hydrogen-electric hybrid system, a fuel cell aging model and a vehicle dynamics model are constructed, and an objective function and corresponding constraint equations at time k are established according to the fuel cell aging model and the vehicle dynamics model;

[0051] The formula of the objective function at time k in step 2) is as follows:

[0052]

[0053] Among them, J(k) represents the total battery aging loss value, N represents the prediction step, χ(i) represents the system optimization variable at the i-th moment, H and f are the quadratic term coefficient matrix and the linear term coefficient matrix respectively, T represents the matrix transpose operation, and i represents the time sequence number; represents the system optimization variable when the total battery aging loss value takes the minimum value;

[0054] The formula of the constraint equation corresponding to the objective function at time k is as follows:

[0055] χ min ≤χ(i)≤χ max

[0056] χ(i) = [X(i), u(i)] T

[0057] X(i) = [IDC (i), V s (i), V f (i), V soc (i), r(i), c(i), θ(i)] T

[0058] X nl (i) = [r(i), c(i), θ(i)] T

[0059] X l (i) = [I DC (i), V s (i), V f (i), V soc (i)] T

[0060] u(i) = ΔI DC (i)

[0061] X l (i + 1) = AX l (i) + B u u(i) + B d d(i)

[0062] X nl (i + 1) = g(X nl (i), u(i))

[0063] Among them, χ min and χ max respectively represent the upper and lower bound matrices of the system state variables. The partial order relation a ≤ b is defined as that the elements at the same positions of a and b both satisfy a i ≤ b i , X(i) represents the system state variable at the i-th moment, u(i) represents the system control variable at the i-th moment, I DC (i) is the output current of the DC / DC at the i-th moment, X nl (i) represents the microscopic state variable related to aging at the i-th moment, V s (i), V f (i) respectively represent the voltages of two capacitors in the lithium battery equivalent circuit model at the i-th moment, V soc (i) represents the percentage unit voltage mapping of the state of charge of the lithium battery at the i-th moment, X l (i) represents the system circuit state variable at the i-th moment, r(i) represents the average radius of platinum particles at the i-th moment, c(i) represents the platinum ion concentration at the i-th moment, θ(i) represents the site ratio at the i-th moment; ΔI DC(i) represents the difference between the output current of the DC converter at the next moment of the i-th moment and the output current of the DC converter at the i-th moment, A, B u 、B d are the linear coefficient matrix, the control variable coefficient matrix, and the disturbance coefficient matrix respectively; X l (i + 1) represents the system circuit state variable at the (i + 1)-th moment, d(i) represents the predicted load current at the i-th moment, that is, the driving condition prediction information within the prediction time domain [k + 1, k + N], X nl (i + 1) represents the microscopic state variable related to aging at the (i + 1)-th moment, g(X nl (i), u(i)) represents the non-linear function of the platinum aging kinetic model at the i-th moment with respect to the microscopic state variable X nl (i) related to aging at the i-th moment and the system control variable.

[0064] χ T (i)Hχ(i) + f T χ(i) = αL fc (i) + βL bat (i), represents the fuel cell aging loss term at the i-th moment, represents the lithium battery aging loss term at the i-th moment, SoC m represents the stable reference value of the state of charge, δ m is the maximum deviation, α and β represent the first and second weight coefficients respectively, ECSA(i) represents the electrochemically active area at the i-th moment. In specific implementation, the stable reference value is set to 0.55, the maximum deviation is set to 0.3, the first and second weight coefficients are taken as 0.85 and 3×10 -9 , the prediction step is taken as 5 steps, the current is limited within 0 to 100A, and the current change rate is limited within 10A / s.

[0065] The setting of the objective function is to balance the aging loss caused by platinum degradation and the power demand in the actual operating conditions of the vehicle. When the state of charge of the lithium battery deviates too much from the applicable value, that is, when it is too high or too low, in the event of a sudden situation, it is very likely that the fuel cell will be forced to passively follow the greatly fluctuating working conditions due to the inability to effectively charge and discharge, thus not only unable to complete power tracking, but also accelerating the aging process of the fuel cell.

[0066] The quadratic term coefficient matrix H, the linear term coefficient matrix f, the linear coefficient matrix A, the control variable coefficient matrix B u and the disturbance coefficient matrix B d The calculation formulas are as follows:

[0067]

[0068]

[0069]

[0070] B u = [T 0 0 0] T

[0071]

[0072] where diag( ) represents a diagonal matrix, 0 1×3 represents a zero vector of dimension 1×3, 0 1×5 represents a zero vector of dimension 1×5, SoC m represents the stable reference value of the state of charge, δ m is the maximum deviation, α and β respectively represent the first and second weight coefficients, r(k - 1) represents the average radius of platinum particles at the (k - 1)-th moment, 0 1×6 represents a zero vector of dimension 1×6, C s , C f and C b respectively represent the capacitance values of the slow dynamic capacitance, fast dynamic capacitance and lithium-ion capacitance, T0 is the sampling time, R s , R f and R sd respectively represent the slow dynamic impedance, fast dynamic impedance and internal impedance. In a specific implementation, the slow dynamic impedance, fast dynamic impedance and internal impedance are respectively taken as 0.01 Ω, 0.008 Ω and 0.5 Ω, the internal resistance of the lithium battery is taken as 0.01 Ω, and the sampling time is taken as 1 s.

[0073] The calculation of the constraint g(X nl (i), u(i)) depends on the following platinum aging kinetic model, and the dynamic equation of the radius is expressed as follows:

[0074]

[0075] where M represents the molar mass of platinum, ρ represents the density of platinum, v1 represents the chemical reaction rate of electrochemical dissolution and Ostwald ripening, and v2 represents the chemical reaction rate of hydrolysis to form a platinum oxide film. The calculation methods of v1 and v2 are as follows:

[0076]

[0077]

[0078] where k i represents the reaction rate constant, β i represents the product of the charge transfer coefficient and the number of electrons of the corresponding reaction, ω i represents the surface interaction parameter specific to the adsorbed substance, η idenotes the overpotential, α i denotes the transfer coefficient, n i denotes the number of transferred electrons, ΔE1 denotes the difference between the electrode potential and the equilibrium potential, c H denotes the hydrogen ion concentration, denotes the hydrogen ion reference concentration, θ i denotes the site ratio, denotes the platinum ion reference concentration, F denotes the Faraday constant, R denotes the universal gas constant, T denotes the temperature.

[0079] Among them, the electrode potential required for the calculation in the difference ΔE1 between the electrode potential and the equilibrium potential can be calculated using the single - cell voltage, and the fuel cell voltage satisfies the following constraints:

[0080] V fc = V o -R fc I fc

[0081] V fc denotes the fuel cell voltage, V o denotes the open - circuit voltage of the fuel cell, R fc denotes the ohmic resistance of the fuel cell, I fc denotes the output current of the fuel cell. For the fuel cell current, it can be calculated through the output current of the DC / DC converter. The DC / DC converter model is regarded as a buck - boost converter model with an efficiency near η when there is effective output energy, that is, V B (k)I DC (k)= η(k)V fc (k)I fc (k).

[0082] The dynamic equation of the change in the concentration of platinum ions is expressed as follows:

[0083]

[0084] Among them, ε denotes the porosity, v4 denotes the chemical reaction rate of the chemical dissolution of the platinum oxide film, q L denotes the mass loss precipitated onto the platinum strip. The calculation methods of v4 and q L are as follows:

[0085]

[0086] Among them, K3 denotes the chemical equilibrium constant, D denotes the diffusion coefficient, denotes the platinum ion concentration, L x denotes the length of the diffused platinum strip, h CL denotes the electrode thickness.

[0087] The dynamic equations for the proportions of unstable oxides and passivated oxides are as follows:

[0088]

[0089] Among them, N C# represents carbon defect sites, v5 represents the chemical reaction rate of the hydrolysis of carbon surface defect sites to form unstable carbon surface oxides, v6 represents the chemical reaction rate of the hydrolysis of unstable carbon surface oxides to form passivated surface oxides, v7 represents the chemical reaction rate of the oxidation of carbon surface oxides by adsorbed water, and v8 represents the chemical reaction rate of the oxidation of carbon surface oxides by hydroxyl groups. Their calculation methods are as follows:

[0090]

[0091]

[0092]

[0093]

[0094] The dynamic equations for the proportions of free platinum particles, hydroxyl groups, and oxygen-adsorbed oxides are as follows:

[0095]

[0096] Among them, N Pt represents platinum sites, and v3 represents the chemical reaction rate of the conversion of platinum film oxidized by hydroxyl group-adsorbed oxides to platinum film oxidized by oxygen atom-adsorbed oxides. The calculation method of v3 is as follows:

[0097]

[0098] 3) According to the real-time information of the driving condition, the predicted information of the driving condition, and the constraint equation, use the sequential quadratic programming algorithm to solve the objective function to obtain the output control current signal of the DC converter in the hydrogen-electric hybrid power system. According to the output control current signal of the DC converter, distribute the output currents of the DC converter and the lithium battery. The output current of the DC converter and the output current of the lithium battery together supply power to the load motor to drive the vehicle to travel in the current prediction time domain;

[0099] 4) Collect the real-time information of the driving condition of the vehicle at the k+N moment in the current prediction time domain and generate the predicted information of the driving condition in the next prediction time domain; repeat steps 2) and 3) to realize the rolling update of the output current control signal of the DC converter.

[0100] In specific implementation, the load current, output current, and lithium battery current change as shown in Figure 3 (a), and the state of charge changes as shown in Figure 3As shown in (b) thereof, the change in the electrochemically active area is as Figure 4 shown.

Claims

1. An energy management method for a hydrogen-electric hybrid power system for anti-aging, characterized in that, It includes the following steps: 1) Collect the real-time driving condition information of the vehicle at the current moment k in real time, and use partial least squares method to perform state prediction based on the real-time driving condition information to generate the driving condition prediction information within the prediction time domain [k + 1, k + N]; 2) Construct a fuel cell aging model and a vehicle dynamics model based on the power characteristics of the hydrogen-electric hybrid system, and establish the objective function and the corresponding constraint equations at the k moment according to the fuel cell aging model and the vehicle dynamics model; 3) According to the real-time driving condition information, the driving condition prediction information and the constraint equations, use the sequential quadratic programming algorithm to solve the objective function, obtain the output control current signal of the DC converter in the hydrogen-electric hybrid system, and drive the vehicle to travel in the current prediction time domain according to the output control current signal of the DC converter; 4) Collect the real-time driving condition information of the vehicle at the k + N moment in the current prediction time domain and generate the driving condition prediction information within the next prediction time domain; repeat steps 2) and 3) to realize the rolling update of the output current control signal of the DC converter; The formula of the objective function at the k moment in step 2) is as follows: Among them, J(k) represents the total battery aging loss value, N represents the prediction step length, χ(i) represents the system optimization variable at the i-th moment, H and f are the quadratic coefficient matrix and the linear coefficient matrix respectively, T represents the matrix transpose operation, and i represents the time sequence number; represents the system optimization variable when the total battery aging loss value takes the minimum value; The formula of the constraint equation corresponding to the objective function at the k moment is as follows: χ min ≤X(i)≤X max χ(i) = [X(i), u(i)] T X(i) = [I DC (i), V s (i), V f (i), V soc (i), r(i), c(i), θ(i)] T X nl (i) = [r(i), c(i), θ(i)] T X l (i) = [I DC (i), V s (i), V f (i), V soc (i)] T u(i) = ΔI DC (i) X l (i + 1) = AX l (i) + B u u(i) + B d d(i) X nl (i + 1)= g(X nl (i), u(i)) where, χ min and χ max represent the upper and lower bound matrices of the system state variables respectively, X(i) represents the system state variable at the i-th moment, u(i) represents the system control variable at the i-th moment, I DC (i) is the output current of the DC converter at the i-th moment, X nl (i) represents the aging-related microscopic state variable at the i-th moment, V s (i), V f (i) represent the voltages of the two capacitors in the lithium battery equivalent circuit model at the i-th moment respectively, V soc (i) represents the percentage unit voltage mapping of the state of charge of the lithium battery at the i-th moment, X l (i) represents the system circuit state variable at the i-th moment, r(i) represents the average radius of the platinum particles at the i-th moment, c(i) represents the platinum ion concentration at the i-th moment, θ(i) represents the site ratio at the i-th moment; ΔI DC (i) represents the difference between the output current of the DC converter at the next moment and the output current of the DC converter at the i-th moment at the i-th moment, A, B u 、B d are the linear coefficient matrix, control variable coefficient matrix, and perturbation coefficient matrix respectively; X l (i + 1) represents the system circuit state variable at the (i + 1)-th moment, d(i) represents the predicted load current at the i-th moment, X nl (i + 1) represents the aging-related microscopic state variable at the (i + 1)-th moment, g(X nl (i), u(i)) represents the non-linear function of the platinum aging kinetic model at the i-th moment with respect to the aging-related microscopic state variable X nl (i) and the system control variable u(i).

2. The energy management method of a hydrogen-electric hybrid system for anti-aging according to claim 1, characterized in that The real-time driving condition information in step 1) includes the voltage and current of the fuel cell at the k moment, the output voltage and current of the DC converter, the voltage and current of the lithium battery, the state of charge, and the real-time load current.

3. The energy management method of a hydrogen-electric hybrid system for anti-aging according to claim 1, characterized in that In step 1), use partial least squares method to perform state prediction based on the real-time driving condition information to generate the driving condition prediction information within the prediction time domain [k + 1, k + N]. The calculation formula is as follows, and the driving condition prediction information is composed of the predicted load current within the prediction time domain [k + 1, k + N]: Among them, represents the predicted load current at the (k + 1)th moment, represents the predicted load current at the (k + N)th moment, where N represents the prediction step size, and I L (k) represents the real-time load current at the kth moment, L1 represents the first fitting coefficient, L2 represents the second fitting coefficient, and I L (k - m1) represents the real-time load current at the (k - m1)th moment before the kth moment, and T represents the matrix transpose operation.

4. The energy management method of a hydrogen-electric hybrid power system for anti-aging according to claim 1, characterized in that The quadratic coefficient matrix H, the linear coefficient matrix f, the linear coefficient matrix A, the control variable coefficient matrix B u and the perturbation coefficient matrix B d are calculated as follows: B u = [T0 0 0 0] T where diag() represents a diagonal matrix, 0 1×3 represents a zero vector with a dimension of 1×3, 0 1×5 represents a zero vector with a dimension of 1×5, SoC m represents the stable reference value of the state of charge, δ m is the maximum deviation, α and β respectively represent the first and second weight coefficients, r(k - 1) represents the average radius of platinum particles at the (k - 1)-th moment, 0 1×6 represents a zero vector with a dimension of 1×6, C s , C f and C b respectively represent the capacitance values of the slow dynamic capacitance, fast dynamic capacitance, and lithium-ion capacitance, T0 is the sampling time, R s , R f and R sd respectively represent the slow dynamic impedance, fast dynamic impedance, and internal impedance.

Images