CN112231830B - Hybrid power vehicle multi-objective optimization control method based on adaptive equivalent factor - Google Patents

Abstract

The invention discloses a hybrid vehicle multi-objective optimization control method based on adaptive equivalent factors. Establishing a hydrogen consumption model, a fuel cell system aging model and a lithium battery system aging model of the hybrid power vehicle, obtaining the hydrogen consumption and the aging state quantity of the fuel cell system and the lithium battery system, and converting the hydrogen consumption and the aging state quantity into energy consumption cost; establishing a semi-empirical fuel cell system aging model, setting an aging parameter, establishing a relation between the aging parameter and the resistance and limiting current of the fuel cell, solving to obtain the aging parameter, and calculating a self-adaptive equivalent factor by using the aging parameter; and establishing a multi-objective function consisting of energy consumption cost and containing self-adaptive equivalent factors, and solving a minimized objective to obtain optimized power distribution for control. The invention can still reasonably distribute the output power through the designed adaptive equivalent factor when the performance of the fuel cell system is attenuated, thereby realizing multi-objective optimization control and having good real-time performance.

Description

Multi-objective optimal control method for hybrid electric vehicle based on adaptive equivalent factor

technical field

The invention relates to a multi-objective optimization control method for a hybrid electric vehicle. In the case of fuel cell system performance degradation, the multi-objective optimization and lithium battery system state-of-charge are guaranteed by designing an adaptive equivalent factor based on the aging state of the fuel cell system. stability while ensuring the real-time performance of the algorithm.

Background technique

Hybrid vehicles have the advantages of zero emission, high energy conversion efficiency, and long cruising range, which have attracted extensive attention from industry and academia. At present, the energy system of hybrid vehicle is mainly composed of fuel cell system and lithium battery system or super capacitor. It is the focus of most current energy control methods to study how to efficiently distribute the energy supply ratio of the two energy systems and improve the economy of hybrid vehicles on the basis of meeting the power requirements. However, the high cost and short service life of fuel cell systems are one of the main reasons for limiting the popularization and application of hybrid electric vehicles. Therefore, it is also necessary to consider the service life of the fuel cell system and the lithium battery system in the hybrid vehicle in the energy control method. In addition, the frequent starting and stopping of the fuel cell system of the hybrid vehicle will cause its performance degradation, thereby reducing the working efficiency of the fuel cell system. This will inevitably increase the use of the lithium battery system to meet the demanded power, thereby causing fluctuations in the state of charge of the lithium battery system to affect the performance of the hybrid vehicle. Therefore, it is necessary to study the multi-objective optimization to improve the economy and durability of hybrid vehicles and ensure the stable state of charge of lithium battery systems.

SUMMARY OF THE INVENTION

In the case of fuel cell system performance degradation, in order to achieve multi-objective optimization of hybrid vehicle economy, durability and stable state of charge of the lithium battery system, the technical solution of the present invention provides a hybrid vehicle based on an adaptive equivalent factor The multi-objective optimal control method of the present invention is that when the performance of the fuel cell system is degraded, the output power can be reasonably distributed through the designed adaptive equivalent factor to realize the multi-objective optimal control, and the real-time performance of the design algorithm is good.

The technical scheme adopted in the present invention is:

Step 1. Establish the hydrogen consumption model, fuel cell system aging model and lithium battery system aging model of the hybrid vehicle:

The hydrogen consumption model of the hybrid vehicle is mainly composed of the hydrogen consumption of the fuel cell system and the equivalent hydrogen consumption of the lithium battery system; the fuel cell system aging model is a model that describes the aging state of the fuel cell system through the number of starts and output power; the lithium battery The system aging model is a model that describes the aging state of the lithium battery system by the discharge rate and operating temperature.

The hydrogen consumption of the fuel cell system is determined by the output power of the fuel cell system, while the equivalent hydrogen consumption of the lithium battery system depends on the output power and the equivalent factor of the lithium battery system.

The output power and equivalent factor of the fuel cell system and lithium battery system are input to the output of the hydrogen consumption model of the hybrid vehicle, and the hydrogen consumption of the hybrid vehicle is obtained. The aging state of the fuel cell system is obtained, and the discharge rate and battery temperature of the lithium battery system are input to the lithium battery system aging model output to obtain the aging state of the lithium battery system.

Step 2. Obtain the hydrogen consumption of the hybrid vehicle, the aging state of the fuel cell system, and the aging state of the lithium battery system through the model established in step 1. The aging state of the fuel cell system and the aging state of the lithium battery system are both 1-100%. In order to convert the hydrogen consumption of the hybrid vehicle, the aging state of the fuel cell system and the aging state of the lithium battery system into the corresponding energy consumption cost;

Step 3. Establish a semi-empirical fuel cell system aging model, set an aging parameter of the fuel cell system, and establish the relationship between the aging parameter and the resistance and limiting current of the fuel cell. The aging parameter can more accurately reflect the aging state of the fuel cell system. ;Using the unscented Kalman filter algorithm to process the semi-empirical fuel cell system aging model, solve and estimate the aging parameters, and use the covariance matching method to calculate the variance of the process noise and measurement noise in the unscented Kalman filter algorithm to improve the unscented Kalman filter. The accuracy of the algorithm estimates. Finally, the adaptive equivalent factor is calculated with the estimated aging parameters, and the adaptive equivalent factor is used to ensure the stability of the state of charge of the lithium battery system;

The invention designs and constructs the adaptive equivalent factor included in the multi-objective function to reasonably distribute the output power of the fuel cell system and the lithium battery system, ensures that the state of charge of the lithium battery system is stable when the performance of the fuel cell system is degraded, and obtains better performance. Accurate fuel cell system aging status.

Among them, the implementation of the present invention adopts a semi-empirical fuel cell system aging model. Using the semi-empirical fuel cell system aging model, it can be obtained that the resistance and limiting current of the fuel cell vary significantly regardless of static load or dynamic load, which improves the estimation accuracy of the fuel cell system's aging state.

Step 4. Establish a multi-objective function including an adaptive equivalent factor composed of the three energy consumption costs in Step 2. The multi-objective function meets the requirements of vehicle power and limits the range of output power of the fuel cell system and the output power of the lithium battery system. , take the multi-objective function as a convex optimization problem with constrained quadratic programming, and then use the effective set algorithm to minimize the multi-objective function to solve the optimal power distribution for control, realize the multi-objective optimal control, and improve the hybrid The economy and durability of power vehicles and ensuring the stability of the state of charge of the lithium battery system.

The hybrid vehicle is provided with a fuel cell system and a lithium battery system, and the fuel cell system and the lithium battery system are connected together to provide energy for the hybrid vehicle.

The step 2 is specifically:

The aging state quantity Δ _fc of the fuel cell system and the aging state quantity Δ _b of the lithium battery system are calculated by the following formulas:

Among them, N _s is the start and stop times of the fuel cell system, δ _s is the start and stop times coefficient of the fuel cell system, P _fc is the output power of the fuel cell system, P _fc,r is the rated power of the fuel cell system, δ ₀ and α ₀ is the first and second attenuation coefficients of the fuel cell system, T is the running time of the hybrid vehicle, I _bat is the output current of the lithium battery system, and the maximum service period N(c, T _c ) of the lithium battery system depends on the discharge Rate c and operating temperature T _c , Q _n is the nominal capacity of the lithium battery system, t is time.

The step 3 is specifically:

The semi-empirical fuel cell system aging model is expressed as:

where N _fc is the number of fuel cells in the fuel cell system, E _o is the open circuit voltage of the fuel cell, I _fc and V _fc are the current and voltage of the fuel cell system, respectively, _{At and B c are} the Tafel constant and Concentration constant, T _o and I _o are the operating temperature and exchange current of the fuel cell, respectively, P _fc and I _l are the resistance and limiting current of the fuel cell, respectively;

The resistance and limiting current of the fuel cell are closely related to the aging state of the fuel cell system.

Then set the aging parameter α of the fuel cell system, and use the aging parameter α to describe the changes of the fuel cell resistance R _fc and the limiting current I _l :

R _fc =R _fco ·(1-α)

I _l =I _lo ·(1+α)

where R _fco and I _lo are the initial fuel cell resistance and limiting current, respectively, and β is the rate of change of the fuel cell system aging parameter α;

Then the unscented Kalman filter algorithm is used to process the semi-empirical fuel cell system aging model to estimate the aging parameters of the fuel cell system, and the covariance matching method is used to update the variance of the process noise and measurement noise in the unscented Kalman filter algorithm in real time;

Finally, the adaptive equivalent factor λ _e is obtained by processing the obtained fuel cell system aging parameters according to the following formula:

λ _e =λ _eo ·(1+k _b α) ²

Among them, λ _eo is the initial equivalent factor, and k _b represents a constant, which is used to adjust the change of the adaptive equivalent factor.

The step 4 is specifically:

Building a multi-objective function is expressed as:

Among them, J _e is the multi-objective function, P _bat is the output power of the lithium battery system, A _f and B _f are the quadratic term coefficient and primary term coefficient of the output power of the fuel cell system, respectively; A _b and B _b are the lithium battery system, respectively Output power quadratic term coefficient and linear term coefficient; wherein the adaptive equivalent factor λ _e is included in the quadratic term coefficient B _b .

为氢气低热值，sgn()为符号函数；Among them, c _bat is the consumption cost per unit of energy of the lithium battery system, E _b,r is the rated capacity of the lithium battery system, A _tol is the total discharge capacity, V _ob is the open circuit voltage of the lithium battery system,

is the low calorific value of hydrogen, and sgn() is the sign function;

The constraints of the multi-objective function are expressed as:

P _fc +P _bat =P _dem

P _fc,min ≤P _fc ≤P _fc,max

P _b,min ≤P _bat ≤P _b,max

Among them, P _dem is the required power of the hybrid vehicle, P _fc,min and P _fc,max are the minimum and maximum output power of the fuel cell system, respectively, P _b,min and P _b,max are the minimum and maximum output power of the lithium battery system, respectively Output Power;

Finally, the multi-objective function is regarded as a constrained quadratic programming problem, and the effective set algorithm is used to solve the optimization, and obtain the optimal output power P _fc of the fuel cell system and the output power P _bat of the lithium battery system, which can ensure the real-time performance of the algorithm. .

The invention can design an adaptive equivalent factor based on the unscented Kalman filter algorithm, reasonably and efficiently distribute the output power of the fuel cell system and the lithium battery system under the condition of the performance degradation of the fuel cell system, and realize multi-objective optimization.

The beneficial effects of the present invention are:

The present invention can ensure the stability of the state of charge of the lithium battery system while improving the hybrid power economy and durability. It can effectively improve the utilization rate of energy and the performance of the lithium battery system, and at the same time increase the service life of the fuel cell system and the lithium battery system.

Description of drawings

Figure 1 is a block diagram of the multi-objective optimization algorithm method;

Figure 2 shows the UDDS driving conditions;

Figure 3 shows the HWFET driving conditions;

Figure 4 Changes in the output power of the fuel cell system under UDDS conditions

Figure 5 Changes in output power of lithium battery system under UDDS conditions

Figure 6. Output power change of fuel cell system under HWFET condition

Figure 7 Output power change of lithium battery system under HFET condition

Figure 8 shows the change of the state of charge of the lithium battery system under UDDS conditions;

Figure 9 shows the change of the state of charge of the lithium battery system under the HWFET operating condition;

Table 1 shows the comparison of the consumption cost of each method under UDDS conditions;

Table 2 shows the comparison of the consumption cost of each method under the HWFET working condition.

Detailed ways

The present invention will be further described below in conjunction with specific examples. It should be understood that these examples are only used to illustrate the present invention and not to limit the scope of the present invention.

The present invention is embodied in a simulation environment to verify the effectiveness of a multi-objective optimal control method for an aging fuel cell system. The specific embodiment and its implementation process are as follows:

1 is a system block diagram of a multi-objective optimal control method provided by the present invention. The workflow of the multi-objective optimal control method is as follows: the required power can be calculated based on the vehicle speed of the hybrid vehicle and the dynamic model of the vehicle. Through the established hybrid vehicle hydrogen consumption model, fuel cell system aging model and lithium battery system aging model, the indicators to measure the economy and durability can be obtained as the total cost. The calculation expression is as follows

C=C _e +C _fc +C _bat

Among them, C _e is the hydrogen consumption cost, which is composed of the hydrogen consumption of the fuel cell system and the equivalent hydrogen consumption of the lithium battery system. The equivalent hydrogen consumption of the lithium battery system is the equivalent of the electric energy consumed by using the adaptive equivalent factor It is converted into hydrogen consumption; C _fc is the loss cost of fuel cell system aging, and C _bat is the loss cost of lithium battery system aging.

The economy and durability of hybrid vehicles are then improved by minimizing a multi-objective function, which is expressed as:

为氢气消耗代价变化率，

为燃料电池系统老化消耗代价变化率，

为燃料电池系统老化消耗代价变化率。多目标函数由能量消耗代价的变化率组成的目的在于能够将多目标优化最终转换为二次规划问题，提高多目标优化控制方法的实时性。in,

is the rate of change of hydrogen consumption cost,

is the rate of change of the aging consumption cost of the fuel cell system,

Rate of change in consumption penalty for fuel cell system aging. The purpose of the multi-objective function consisting of the rate of change of the energy consumption cost is to convert the multi-objective optimization into a quadratic programming problem and improve the real-time performance of the multi-objective optimization control method.

In order to ensure the stability of the state of charge of the lithium battery system under the performance degradation of the fuel cell system, based on the semi-empirical fuel cell system aging model, the unscented Kalman filter algorithm is used to accurately estimate the aging parameters of the fuel cell system, so as to obtain the adaptive equivalent The factor maintains the stable state of charge of the lithium battery system.

The above semi-empirical fuel cell system aging model is expressed as:

where N _fc is the number of fuel cells in the fuel cell system, E _o is the open circuit voltage of the fuel cell, I _fc and V _fc are the current and voltage of the fuel cell system, respectively, _{At and B c are} the Tafel constant and The concentration constants, T _o and I _o are the operating temperature and exchange current of the fuel cell, respectively.

The resistance R _fc and limiting current I _l of the fuel cell are closely related to the aging state of the fuel cell system.

Then set the aging parameter α of the fuel cell system, and use the aging parameter α to describe the changes of the fuel cell resistance R _fc and the limiting current I _l :

R _fc =R _fco ·(1-a)

I _l =I _lo ·(1+α)

Among them, R _fco and I _lo are the initial fuel cell resistance and limiting current, respectively, and β is the rate of change of the fuel cell system aging parameter α. Then, the unscented Kalman filter algorithm is used to estimate the aging parameters of the fuel cell system, and the covariance matching method is used to update the variance of the process noise and measurement noise in the unscented Kalman filter algorithm in real time. Finally, based on the estimated fuel cell system aging parameters, the adaptive equivalent factor λ _e is expressed as:

λ _e =λ _eo ·(1+k _b α) ²

Among them, λ _eo is the initial equivalent factor, and the constant k _b is used to adjust the change of the adaptive equivalent factor. Then simplify the multi-objective function to get the following expression:

Among them, P _bat is the output power of the lithium battery system, A _f and B _f are the quadratic term coefficient and primary term coefficient of the output power of the fuel cell system, respectively; A _b and B _b are the quadratic term coefficient of the output power of the lithium battery system and The first-order coefficients; where the adaptive equivalent factor λ _e is included in the quadratic coefficients B _b .

为氢气低热值。Among them, c _bat is the consumption cost per unit of energy of the lithium battery system, E _b,r is the rated capacity of the lithium battery system, A _tol is the total discharge capacity, V _ob is the open circuit voltage of the lithium battery system,

Low calorific value for hydrogen.

In order to meet the power requirements of the vehicle and ensure that the output power of the fuel cell system and the lithium battery system is limited to the allowable variation range, the equation and inequality constraints of the multi-objective function are established, which are expressed as:

P _fc +P _bat =P _dem

P _fc,min ≤P _fc ≤P _fc,max

P _b,min ≤P _bat ≤P _b,max

Among them, P _fc,min and P _fc,max are the minimum and maximum output power of the fuel cell system, respectively, and P _b,min and P _b,max are the minimum and maximum output power of the lithium battery system, respectively.

Finally, the multi-objective function is regarded as a constrained quadratic programming problem, and the effective set algorithm is used to solve the optimization, and obtain the optimal output power P _fc of the fuel cell system and the output power P _bat of the lithium battery system, and can ensure the real-time algorithm. performance.

The method of the invention is verified on a simulation platform, and the algorithm is verified under two standard operating conditions. The initial state of charge of the lithium battery system is set to 0.7, and the initial equivalence factor is determined according to the new fuel cell system. The maximum output power of the fuel cell system and lithium battery system is 50kW and 25kW, respectively. Then compared the simulation results of the energy control method based on the constant equivalent factor and the adaptive equivalent factor when the fuel cell system is aging, the difference between the two energy control optimization methods is only whether to update the equivalent factor. At the same time, the commonly used minimum equivalent energy consumption method is proposed to compare the effectiveness of the designed multi-objective energy control method. The operating conditions of the two simulations are shown in Figures 2 and 3. The power distribution under the two working conditions is shown in Figure 4-Figure 7. Figure 8 and Figure 9 show the changes in the state of charge of the battery under the two working conditions. In addition, the comparison of the energy control consumption cost of each method under the two working conditions is shown in Table 1 and Table 2. It can be seen from the simulation results that although the method of minimizing the equivalent energy consumption reduces the aging loss of the lithium battery system, it increases the aging loss of the fuel cell system. Compared with the multi-objective optimal control method, the method with the smallest equivalent energy consumption has the highest total energy consumption cost. Compared with the multi-objective optimal control method of the constant equivalent factor, the adaptive equivalent factor can maintain the state of charge of the lithium battery system around the initial value.

Table 1 Consumption cost comparison of various methods under UDDS conditions

method Minimum equivalent energy consumption Constant Equivalent Factor Adaptive Equivalence Factor C_e($) 0.301 0.2625 0.229 C_bat($) 6.909e-08 1.673e-08 1.427e-07 C_fc($) 1.672 1.65 1.641 total 1.973 1.913 1.87

Table 2 Consumption cost comparison of various methods under HWFET conditions

method Minimum equivalent energy consumption Constant Equivalent Factor Adaptive Equivalence Factor C_e($) 0.3316 0.2771 0.2462 C_bat($) 4.678e-08 1.397e-07 1.164e-07 C_fc($) 0.7694 0.7549 0.7469 total 1.101 1.032 0.9931

Therefore, compared with other control methods, the method of the present invention has the lowest total consumption cost under different working conditions, which means that the designed multi-objective optimal control method can balance the economy and durability of hybrid vehicles, and at the same time, the method can also maintain the lithium battery. Stability of the state of charge of the system.

Claims (3)

1. A hybrid vehicle multi-objective optimal control method based on an adaptive equivalent factor is characterized in that: Step 1. Establish the hydrogen consumption model, fuel cell system aging model and lithium battery system aging model of the hybrid vehicle: Step 2. Obtain the hydrogen consumption of the hybrid vehicle, the aging state of the fuel cell system and the lithium battery system through the model established in step 1, so as to combine the hydrogen consumption of the hybrid vehicle, the aging state of the fuel cell system and the lithium battery. The aging state of the battery system is converted into the corresponding energy consumption cost; Step 3. Establish a semi-empirical fuel cell system aging model, set an aging parameter of the fuel cell system, establish the relationship between the aging parameter and the fuel cell resistance and limited current, and use the unscented Kalman filter algorithm to process the semi-empirical fuel cell system aging. The model is solved and estimated to obtain the aging parameters, and the variance of the process noise and the measurement noise in the unscented Kalman filter algorithm is calculated by the covariance matching method, and finally the adaptive equivalent factor is calculated with the estimated aging parameters; Step 4, establishing a multi-objective function including an adaptive equivalent factor composed of the three energy consumption costs in Step 2, and then using the effective set algorithm to minimize the multi-objective function as the goal to solve to obtain an optimized power distribution for control; The step 3 is specifically: The semi-empirical fuel cell system aging model is expressed as:

where N _fc is the number of fuel cells in the fuel cell system, E _o is the open circuit voltage of the fuel cell, I _fc and V _fc are the current and voltage of the fuel cell system, respectively, _{At and B c are} the Tafel constant and Concentration constants, _{To and I o are} the operating temperature and exchange current of the fuel cell, respectively, and R _fc and I _l represent the resistance and limiting current of the fuel cell, respectively; Then set the aging parameter α of the fuel cell system, and use the aging parameter α to describe the changes of the fuel cell resistance R _fc and the limiting current I _l : R _fc =R _fco ·(1-α) I _l =I _lo ·(1+α)

where R _fco and I _lo are the initial fuel cell resistance and limiting current, respectively, and β is the rate of change of the fuel cell system aging parameter α; Then the unscented Kalman filter algorithm is used to process the semi-empirical fuel cell system aging model to estimate the aging parameters of the fuel cell system, and the covariance matching method is used to update the variance of the process noise and measurement noise in the unscented Kalman filter algorithm in real time; Finally, the adaptive equivalent factor λ _e is obtained by processing the obtained fuel cell system aging parameters according to the following formula: λ _e =λ _eo ·(1+k _b α) ² Among them, λ _eo is the initial equivalent factor, and k _b is a positive constant; The step 4 is specifically: Building a multi-objective function is expressed as:

Among them, J _e is the multi-objective function, P _bat is the output power of the lithium battery system, A _f and B _f are the quadratic term coefficient and primary term coefficient of the output power of the fuel cell system, respectively; A _b and B _b are the lithium battery system, respectively Output power quadratic term coefficient and primary term coefficient;

Among them, c _bat is the consumption cost per unit of energy of the lithium battery system, E _b,r is the rated capacity of the lithium battery system, A _tol is the total discharge capacity, V _ob is the open circuit voltage of the lithium battery system,

is the low calorific value of hydrogen, _sgn () is the sign function, and P _fc is the output power of the fuel cell system; The constraints of the multi-objective function are expressed as: P _fc +P _bat =P _dem P _fc,min ≤P _fc ≤P _fc,max P _b,min ≤P _bat ≤P _b,max Among them, P _dem is the required power of the hybrid vehicle, P _fc,min and P _fc,max are the minimum and maximum output power of the fuel cell system, respectively, P _b,min and P _b,max are the minimum and maximum output power of the lithium battery system, respectively Output Power; Finally, the multi-objective function is regarded as a constrained quadratic programming problem, and the effective set algorithm is used to solve the optimization, and obtain the optimal output power P _fc of the fuel cell system and the output power of the lithium battery system. 2 . The multi-objective optimal control method for a hybrid electric vehicle based on an adaptive equivalent factor according to claim 1 , wherein the hybrid electric vehicle is provided with a fuel cell system and a lithium battery system, and the fuel cell The system and the lithium battery system are connected together to provide energy for the hybrid vehicle. 3. A hybrid vehicle multi-objective optimization control method based on an adaptive equivalent factor according to claim 1, wherein the step 2 is specifically: The aging state quantity Δ _fc of the fuel cell system and the aging state quantity Δ _b of the lithium battery system are calculated by the following formulas:

Among them, N _s is the start-stop times of the fuel cell system, δ _s is the start-stop times coefficient of the fuel cell system, P _fc,r is the rated power of the fuel cell system, δ ₀ and α ₀ are the first, The second attenuation coefficient, T is the running time of the hybrid vehicle, I _bat is the output current of the lithium battery system, Q _n is the nominal capacity of the lithium battery system, N(c, T _c ) is the maximum service cycle of the lithium battery system, t is time.

Images