A system and method for fuel consumption allocation based on soc prediction

Abstract

The application discloses a fuel consumption distribution system and method based on SOC prediction, comprising: a hybrid vehicle energy management model, which is used to realize the constraint relationship of wheel torque, engine torque and motor torque, and the constraint relationship is that the distribution ratio of engine torque and motor torque is controlled according to the optimal fuel consumption factor; an equivalent circuit model, which is used to simulate the dynamic change of lithium battery during driving, acquire and output the internal parameters of the battery, and establish an observation equation; a lithium battery SOC prediction model, which is used to calculate the optimal predicted SOC value at time t; and a fuel consumption distribution module, which is used to call the hybrid vehicle energy management model to calculate and generate the optimal fuel consumption factor according to the optimal predicted SOC value, so as to realize the distribution of engine torque and motor torque. According to the above technical scheme, the accuracy of the SOC prediction value can be improved, and on this basis, the real-time and accurate optimal fuel consumption factor can be obtained, so that the optimal real-time application of the energy management strategy can be realized.

Description

Technical Field

[0001] This invention relates to the field of energy management strategies for hybrid electric vehicles, and more specifically, to a fuel consumption allocation system and method based on SOC prediction. Background Technology

[0002] Currently, hybrid power remains one of the effective ways to achieve energy conservation and emission reduction in automobiles. Energy management strategies have a crucial impact on realizing this potential, and a well-designed energy management strategy can significantly optimize vehicle economy and emissions performance. Current energy management strategies are mainly divided into two categories: rule-based energy management strategies and optimization-based energy management strategies. Optimization-based control strategies, such as dynamic programming and the equivalent fuel consumption minimization strategy, can achieve optimal fuel economy. Among these, the equivalent fuel consumption minimization strategy (ECMS) is a real-time optimization strategy and is most likely to be applied in real vehicles, but its computational complexity limits its application. Rule-based control strategies are simple and can be applied to real vehicles, but rule formulation mainly relies on human experience, resulting in poor control performance. Furthermore, these methods all have certain drawbacks: for example, they require rules extracted based on one or more specific operating conditions, resulting in weak robustness and generally only applicable to specific operating conditions.

[0003] Therefore, a method is needed to solve for the optimal equivalent fuel consumption factor, which can both achieve real-time prediction to overcome the shortcomings of ECMS in terms of real-time performance, and ensure the accuracy of allocation so that ECMS can play a better role. Summary of the Invention

[0004] In a first aspect, to achieve the above objectives, the present invention provides a fuel consumption allocation system based on SOC prediction, comprising:

[0005] A hybrid electric vehicle energy management model is used to realize the constraint relationship between wheel torque, engine torque and motor torque. The constraint relationship refers to controlling the distribution ratio of engine torque and motor torque according to the optimal fuel consumption factor.

[0006] An equivalent circuit model is used to simulate the dynamic changes of lithium batteries during driving, collect the current and terminal voltage of lithium batteries during driving, obtain and output the internal parameters of the battery, and establish observation equations; wherein, the internal parameters of the battery include terminal current, terminal voltage, open circuit voltage, ohmic resistance, electrochemical polarization resistance, deep polarization resistance, electrochemical polarization capacitance, and deep polarization capacitance.

[0007] A lithium battery SOC prediction model is used to obtain the observation equation and calculate the optimal predicted SOC value at time t based on the observation equation.

[0008] The fuel consumption allocation module is used to calculate and generate the optimal fuel consumption factor based on the best predicted SOC value, calling the hybrid vehicle energy management model to achieve the allocation of engine torque and electric motor torque; the calculation method for the optimal fuel consumption factor is as follows:

[0009] s(t)=s eq0 +ξ(SOC ref -SOC(t)); where s(t) is the real-time equivalent consumption factor, s eq0 For the equivalent fuel factor constant part, SOC ref Let ξ be the initial SOC value and ξ be the penalty factor.

[0010] The lithium battery SOC prediction model includes:

[0011] The EKF prediction model is used to estimate the SOC value and terminal voltage prediction value at time t based on the SOC value of the previous time step, the initial SOC value, and the battery internal parameters.

[0012] The SABO algorithm unit is used to correct measurement errors in battery terminal voltage and current, and to adjust current and terminal voltage.

[0013] The posterior estimation unit is used to perform posterior estimation based on the EKF algorithm. Based on the corrected current, terminal voltage value, prior estimated SOC value at time t, and predicted terminal voltage value, it outputs the optimal SOC value at time t.

[0014] The equivalent circuit model is a vehicle battery model based on the Thevenin model with the addition of a second-order RC equivalent circuit model. The structural elements of the equivalent circuit model include ohmic resistors, electrochemical polarization resistors, deep polarization resistors, electrochemical polarization capacitors, and deep polarization capacitors.

[0015] Furthermore, the equivalent circuit model uses the least squares method to identify parameters, simulating the real-time acquisition of battery terminal current and voltage data during vehicle operation, and online identification and correction of the specific values ​​of battery internal parameters.

[0016] Furthermore, the observation equations established by the equivalent circuit model are generated based on the Laplace transform of the battery's internal parameters, and are expressed as follows:

[0017] Among them, U ocv For open circuit voltage, U out I is the terminal voltage, R0 is the ohmic resistance, C1 is the electrochemically polarized capacitor, R1 is the electrochemically polarized resistor, C2 is the deeply polarized capacitor, R2 is the deeply polarized resistor, and s is a plane in the Laplace transform process.

[0018] On the other hand, the present invention provides a fuel consumption allocation method based on SOC prediction, comprising the following steps:

[0019] Establish a hybrid electric vehicle energy management model, and realize the constraint relationship between wheel torque, engine torque and motor torque in the management model. The constraint relationship refers to: realizing the distribution ratio control of engine torque and motor torque according to the optimal fuel consumption factor.

[0020] An equivalent circuit model is established to simulate the dynamic changes of the lithium battery during driving. The current and terminal voltage of the lithium battery during driving are collected to obtain the internal parameters of the battery and establish observation equations. The internal parameters of the battery include terminal current, terminal voltage, turn-on voltage, ohmic resistance, electrochemical polarization resistance, deep polarization resistance, electrochemical polarization capacitance, and deep polarization capacitance.

[0021] A lithium battery SOC prediction model is established to predict the optimal SOC value at time t based on the observation equation.

[0022] Based on the optimal predicted SOC value, the hybrid vehicle energy management model is invoked to calculate and generate the optimal fuel consumption factor, thereby achieving the distribution of engine torque and electric motor torque; wherein, the calculation method of the optimal fuel consumption factor is as follows:

[0023] s(t)=s eq0 +ξ(SOC ref -SOC(t)); where s(t) is the real-time equivalent consumption factor, s eq0 For the equivalent fuel factor constant part, SOC ref Let ξ be the initial SOC value and ξ be the penalty factor.

[0024] The steps involved in establishing a lithium battery SOC prediction model are as follows:

[0025] An EKF prediction model is established. Based on the SOC value of the previous time step, the initial SOC value, and the current measurement value, the EKF prediction model estimates the SOC value and the predicted terminal voltage value at time t.

[0026] The SABO algorithm is incorporated to correct measurement errors in battery terminal voltage and current.

[0027] Based on measurement error correction, prior estimation of the SOC value at time t, and predicted terminal voltage, the optimal SOC value at time t is output.

[0028] The equivalent circuit model is a vehicle battery model based on the Thevenin model with the addition of a second-order RC equivalent circuit model. The elements in the equivalent circuit model structure include ohmic resistance, electrochemical polarization resistance, deep polarization resistance, electrochemical polarization capacitance, and deep polarization capacitance.

[0029] Furthermore, the least squares method is used to identify the parameters in the equivalent circuit model, and the battery terminal current and terminal voltage data can be collected in real time during the simulation of the car driving process, so as to identify and correct the internal parameters of the battery online.

[0030] The observation equation is generated based on the Laplace transform of the battery's internal parameters and is expressed as follows:

[0031] Among them, U ocv For open circuit voltage, U out I is the terminal voltage, R0 is the ohmic resistance, C1 is the electrochemically polarized capacitor, R1 is the electrochemically polarized resistor, C2 is the deeply polarized capacitor, R2 is the deeply polarized resistor, and s is a plane in the Laplace transform process.

[0032] According to the present invention, the accuracy of the optimal SOC prediction value can be improved, and on this basis, the optimal fuel consumption factor can be obtained in real time and accurately, thereby realizing the optimal real-time application of energy management strategies. Attached Figure Description

[0033] Figure 1 This is a schematic diagram of a fuel consumption distribution system based on SOC prediction provided in an embodiment of the present invention;

[0034] Figure 2 This is a schematic diagram of the ECMS process provided according to an embodiment of the present invention;

[0035] Figure 3 This is a schematic diagram of the lithium battery SOC prediction process provided in an embodiment of the present invention;

[0036] Figure 4 This is a schematic diagram of the existing lithium battery SOC prediction process;

[0037] Figure 5 This is a structural diagram of a hybrid electric vehicle's powertrain system;

[0038] Figure 6 This is a schematic diagram of an equivalent circuit model RC circuit provided according to an embodiment of the present invention;

[0039] Figure 7 This is a flowchart illustrating the steps of a fuel consumption allocation method provided in an embodiment of the present invention. Detailed Implementation

[0040] The solution of this invention is applied to hybrid electric vehicles and is suitable for powertrain structures such as... Figure 5The hybrid powertrain system shown primarily comprises an engine, electric motor, clutch, transmission, and battery. The electric motor is directly connected to the clutch output and the transmission input, with the motor rotor also functioning as a torque coupler. By omitting the mechanical coupler, this powertrain system boasts a simple and compact structure. In this configuration, the electric motor functions as an engine starter, a generator in regenerative braking, and an engine auxiliary. This vehicle model can freely switch between four modes: engine-only drive, motor-only drive, hybrid drive, and regenerative braking. Based on this structure, an energy management strategy for plug-in hybrid electric vehicles needs to be developed based on the theory of minimum fuel consumption.

[0041] The solution provided by this invention employs the Equivalent Fuel Minimum Strategy (ECMS) for energy management. Addressing the real-time variation of the equivalent fuel factor in ECMS, this paper uses battery charge balancing as a constraint and the SABO-EKF algorithm to predict the real-time battery SOC sequence. Using this optimal SOC sequence as a reference trajectory, the corresponding equivalent factor is solved in real-time during actual vehicle operation as the optimal fuel consumption factor for allocating engine torque and motor torque, thereby achieving real-time application of the energy management strategy.

[0042] The specific implementation of the present invention will now be described in detail with reference to the accompanying drawings.

[0043] Figure 1 A schematic diagram of the fuel consumption distribution system provided by this invention is provided, as shown in the figure, including: a P100 hybrid vehicle energy management model, a P110 equivalent circuit model, a P120 lithium battery SOC prediction model, and a P130 fuel consumption distribution module.

[0044] The P100 hybrid vehicle energy management model is used to realize the constraint relationship between wheel torque, engine torque and motor torque. The constraint relationship refers to obtaining the distribution ratio of engine torque and motor torque according to the optimal fuel consumption factor.

[0045] Since the energy management strategy design is positively correlated with the vehicle's longitudinal speed but independent of its speeds in other directions, this invention ignores the vehicle's lateral dynamics and driving stability, modeling vehicle motion solely based on longitudinal dynamics. According to the dynamic equilibrium relationship during vehicle operation, the vehicle's longitudinal dynamics formula is:

[0046] Wherein: T w R is the wheel torque; w Where is the tire radius; G is gravity; f is the tire rolling resistance coefficient; α is the slope angle; C D U is the air drag coefficient; A is the frontal area; u aδ is the vehicle speed; m is the mass conversion factor; δ is the total mass of the vehicle.

[0047] The relationship between wheel torque, engine torque, and motor torque is as follows:

[0048] T w =(T eng +T mot )i g i0η T +T b , where i g For the transmission speed ratio, i0 is the main transmission ratio, and η is the transmission speed ratio. T For transmission efficiency, T eng For engine torque, T mot Motor torque, T b This is the braking torque.

[0049] Based on the above principles, the equivalent fuel consumption strategy uses the equivalent fuel factor to convert electricity consumption into equivalent fuel consumption. Then, the sum of engine fuel consumption and motor consumption is taken as the total instantaneous equivalent fuel consumption of the vehicle. Finally, the optimal solution in the control domain is solved using the minimum principle.

[0050] The total instantaneous equivalent fuel consumption of the vehicle is: Among them, s eq It is the equivalent fuel consumption factor; This refers to the vehicle's instantaneous equivalent fuel consumption. This refers to the engine's steady-state fuel consumption. P is the equivalent fuel consumption of the motor; Q is the instantaneous power of the motor; and Q is the calorific value constant of the fuel.

[0051] The hybrid vehicle energy management model, applied to hybrid vehicles, achieves an equivalent fuel minimization strategy (ECMS) through the control of the equivalent fuel consumption factor. Its overall logical structure is as follows: Figure 2 As shown.

[0052] The P210 section describes the powertrain structure of hybrid vehicles. This section dynamically outputs vehicle speed, changes in lithium battery dynamics, and drive power demand to obtain the optimized engine torque T. eng Motor torque T mot This enables control over fuel distribution during the operation of hybrid vehicles.

[0053] The calculation of the equivalent factor in its constraint relationship, namely the optimal fuel consumption factor, is implemented in section P220: based on the lithium battery voltage and current information output by the power system, the SOC value is predicted, and the equivalent factor is solved in real time; the constraint relationship in the hybrid vehicle energy management model is realized through the optimal fuel consumption factor, such as... Figure 2See section P201 in the document. In this invention, during the process of predicting the SOC value, the commonly used SOC prediction model and the observation equation input into the model are optimized to obtain the optimal fuel consumption factor.

[0054] The P110 equivalent circuit model is used to simulate the dynamic changes of lithium batteries during driving. It collects the current and terminal voltage of lithium batteries during driving, obtains the internal parameters of the battery, and establishes observation equations. The internal parameters of the battery include terminal current, terminal voltage, open circuit voltage, ohmic resistance, electrochemical polarization resistance, deep polarization resistance, electrochemical polarization capacitance, and deep polarization capacitance.

[0055] A reasonable equivalent circuit model should be able to accurately describe the dynamic changes of a simulated lithium battery, and should also have easily identifiable model parameters and be readily applicable in practical engineering. Among equivalent circuit models, the Thevenin model stands out for its advantages; it is not only computationally simple but also reflects the battery's dynamic performance to a certain extent through the RC circuit. Therefore, to fully consider the trade-off between model accuracy and computational complexity, the equivalent circuit model provided in this invention adds an RC circuit to the Thevenin model, establishing a second-order RC equivalent circuit model for the vehicle battery model, as shown in the figure below. Figure 6 As shown, where U ocv with U out These are the open-circuit voltage and the terminal voltage, respectively.

[0056] Based on the circuit model, the equivalent circuit model of the second-order RC circuit is as follows:

[0057]

[0058] The parameters in the formula are the internal parameters of the battery, including resistance and capacitance; resistance further includes ohmic resistance, electrochemical polarization resistance, and deep polarization resistance, and capacitance further includes electrochemical polarization capacitance and deep polarization capacitance.

[0059] and Figure 2 The model shown corresponds to the following relationship: I represents current, R0 represents ohmic resistance, C1 represents electrochemical polarization capacitance, R1 represents electrochemical polarization resistance, C2 represents deep polarization capacitance, R2 represents deep polarization resistance, and U1 and U2 represent electrochemical electrode and deep electrode voltages, respectively. The derivatives of U1 and U2;

[0060] Discretizing the above equation and performing a Laplace transform on the terminal voltage, we can then obtain the observation equation as follows:

[0061]

[0062] Among them, U ocv For open circuit voltage, U outI is the terminal voltage, R0 is the ohmic resistance, C1 is the electrochemically polarized capacitor, R1 is the electrochemically polarized resistor, C2 is the deeply polarized capacitor, R2 is the deeply polarized resistor, and s is a plane in the Laplace transform process.

[0063] Meanwhile, the least squares method is used to identify parameters in the equivalent circuit model. During the simulation of a car driving, the battery's terminal current and voltage data can be collected in real time. Based on the circuit structure in the equivalent circuit model, the internal parameters of the battery are identified online, and their specific values ​​are corrected online. This effectively alleviates the problem that new data is difficult to correct in the later stages of the model due to the increase in data and correction pressure.

[0064] The P120 lithium battery SOC prediction model is used to obtain the observation equation and, based on the observation equation, to predict the optimal predicted SOC value at time t.

[0065] In existing technologies, the EKF prediction model is often used to predict the SOC value, such as... Figure 4 As shown in the figure, the input terminal current and terminal voltage are used in the model. The ampere-hour integral method is used as the state method. The spatial model is established based on the observation equation output by the battery model. Based on the SOC value of the previous moment or the initial SOC value and the terminal current value, the SOC value and the terminal voltage prediction value are estimated a priori. Then, based on the parameter relationship, the error of the voltage observation value and the prediction value is calculated. Finally, the gain matrix is ​​calculated based on the a priori estimate to estimate the SOC value posteriorly.

[0066] The lithium battery SOC prediction model in this invention is based on the EKF prediction model. Furthermore, the lithium battery SOC prediction model includes: P121 EKF prediction model, P122 SABO algorithm unit, and P123 posterior estimation unit.

[0067] P121: EKF prediction model, used to estimate the SOC value and terminal voltage prediction value at time t based on the SOC value of the previous time step, the initial SOC value and the battery internal parameters.

[0068] P122: SABO algorithm unit, used to correct measurement errors in battery terminal voltage and current, and to adjust current and terminal voltage.

[0069] P123: Posterior estimation unit, used to perform posterior estimation based on EKF algorithm, and output the optimal SOC value at time t based on the measurement error correction, prior estimation of SOC value at time t and terminal voltage prediction value.

[0070] The process of lithium battery SOC prediction model is as follows: Figure 3 As shown:

[0071] Step S301: First, collect the terminal current and terminal voltage values ​​from the lithium battery;

[0072] Steps S331 and S332: Using the equivalent circuit model, the least squares algorithm is used to identify parameters. Based on the terminal current and terminal voltage, various circuit parameters (capacitance and resistance) are added to establish the state equation and observation equation.

[0073] In the subsequent implementation of the EKF algorithm, the SABO algorithm was incorporated to modify the measurement error, specifically including the following steps:

[0074] Steps S311 and S312: First, establish the EKF prediction model and perform SOC prior estimation using the EKF algorithm:

[0075] Using the ampere-hour integral method as the state equation, the state-space model can be obtained by simultaneously solving the circuit model and the observation equation, as follows:

[0076]

[0077] f(x) in the formula k ,u k ) and g(x k ,u k ) are the state equation and the observation equation, respectively; and g(x) k ,u k Specifically, it is expressed as

[0078] At any given moment, performing a first-order Taylor expansion on the state equation and the observation equation yields the coefficients A in the extended Kalman filter algorithm. k and C k for:

[0079]

[0080]

[0081] Substituting the above system into the state-space model and linearizing it, we obtain the following state-space model:

[0082]

[0083] Using this state-space model, the following calculations are performed:

[0084] Step 1: Obtain the initial SOC value and the initial covariance matrix P 0|0 , is represented as:

[0085]

[0086] Step 2: Solve the state equation A k and C k ;

[0087] Step 3: Predict future states to obtain prior estimates and Represented as: y k|k-1 =g(x k|k-1 ,u k )+v k ;

[0088] Step 4: Prediction error covariance, expressed as:

[0089] Q k-1 These are the parameters for the covariance formula;

[0090] Step 5: Calculate the Kalman filter gain K k , is represented as:

[0091] Where R k The noise variance is measured, and its magnitude directly affects the correction effect of the EKF algorithm.

[0092] That is, the prediction model is based on the SOC value of the previous time step. Alternatively, the initial SOC value and the current measurement value can be used to estimate the SOC value at time t a priori. Sum terminal voltage prediction value

[0093] Steps S321 to S324: Add the SABO algorithm to correct measurement errors in battery terminal voltage and current;

[0094] Because errors are unavoidable in measuring the terminal voltage and current of lithium batteries, and because the established equivalent circuit model simulates the complex dynamic model of the battery, the identification of internal battery parameters will further exacerbate the deviation on top of the existing errors. Therefore, in this invention, an optimization algorithm based on subtraction averaging is established to correct the noise variance R. k Q k The value of .

[0095] The steps of the SABO algorithm are as follows:

[0096] Step 1: Initialize the population:

[0097] x i,j =lb j +r(ub j -lb j ), where x i,j For SOC individuals, lb j ub represents the lower bound of the j-th dimension variable. j This represents the upper bound of the j-th dimension variable, where r is a random number between [0,1].

[0098] Step 2: SABO search process:

[0099] Where F(A) and F(B) are the fitness values ​​of individuals A and B, respectively. It is a vector of randomly generated data in {0,1};

[0100] The iterative search process is represented as:

[0101]

[0102] Where N is the total number of individuals. The updated position is used to replace the original position if the updated position is better; otherwise, the original position is retained.

[0103] By combining optimization of the EKF noise variance matrix, the optimization matrix Q can be achieved. k and R k The optimal SOC value is determined by the following algorithm:

[0104]

[0105] Furthermore, the predicted value of the terminal voltage is selected. With the observed value U k The absolute differences are accumulated and used as the fitness value, that is:

[0106] In the formula: L represents the maximum number of sampling points at discrete frequency points;

[0107] Steps S313 and S314: The EKF algorithm performs posterior estimation, and based on the measurement error correction, the prior estimate of the SOC value at time t, and the predicted terminal voltage value, outputs the optimal SOC value at time t.

[0108] The optimal noise variance R is obtained through iterative SABO algorithm. k+1 Q k+1 Calculate the Kalman gain based on the value of , and update the error covariance, i.e.:

[0109]

[0110]

[0111] Step S314: Calculate the optimal estimate of SOC. The algorithm is expressed as follows:

[0112] Where, x k This is the optimal predicted value for SOC.

[0113] Based on the optimal prediction value for SOC, it is possible to achieve Figure 2The equivalent factor of the S220 part is solved in real time, and the optimal fuel consumption factor is obtained.

[0114] P130: Fuel consumption allocation module, used to calculate and generate the optimal fuel consumption factor based on the best predicted SOC value at time t, by calling the hybrid vehicle energy management model, to realize the allocation of engine torque and motor torque; the selection of the equivalent fuel consumption factor has a significant impact on the performance of the control strategy. In this step, by solving for a suitable equivalent fuel factor in real time, the potential of ECMS is fully utilized to achieve the best fuel economy for the whole vehicle.

[0115] In the specific calculations, the optimization objective is to achieve optimal fuel economy, and the performance functional is:

[0116] Where SOC(t) is the optimal SOC value prediction at time t, which can be expressed as x output by the P120 lithium battery SOC prediction model. k .

[0117] To maintain SOC balance, the optimization objective functional can be defined as follows:

[0118]

[0119] Among them, SOC ref ξ is the SOC reference value, and ξ is the penalty factor. For overall efficiency, Q batt Q is the calorific value constant of the battery. lhv is the calorific value constant of fuel oil.

[0120] The objective function consists of three parts: the first part is the engine fuel consumption, which is independent of the battery SOC; the second part is the equivalent fuel consumption of the electric motor; and the third part is the penalty function for maintaining the battery SOC balance.

[0121] Therefore, the method for calculating the optimal fuel consumption factor is as follows:

[0122] s(t)=s eq0 +ξ(SOC ref -SOC(t));

[0123] Where s(t) is the real-time equivalent consumption factor, s eq0 For the equivalent fuel factor constant part, SOC ref Let ξ be the initial SOC value and ξ be the penalty factor.

[0124] like Figure 2 As shown, in section S200, based on the real-time equivalent consumption factor, the engine fuel consumption, the electric motor equivalent fuel consumption, and the corresponding electric motor and engine power are calculated to realize the engine torque T. eng Motor torque Tmot Through optimization, the real-time application of energy management strategies was ultimately achieved.

[0125] Figure 7 A fuel consumption allocation method based on SOC prediction is provided, including the following steps:

[0126] Step S700: Establish a hybrid vehicle energy management model, and realize the constraint relationship between wheel torque, engine torque and motor torque in the management model. The constraint relationship mentioned in this invention refers to: realizing the distribution ratio control of engine torque and motor torque according to the optimal fuel consumption factor.

[0127] Step S710: Establish an equivalent circuit model to simulate the dynamic changes of the lithium battery during driving, collect the current and terminal voltage of the lithium battery during driving, obtain the internal parameters of the battery, and establish observation equations. The internal parameters of the battery include terminal current, terminal voltage, turn-on voltage, ohmic resistance, electrochemical polarization resistance, deep polarization resistance, electrochemical polarization capacitance, and deep polarization capacitance.

[0128] The equivalent circuit model is a vehicle battery model based on the Thevenin model with the addition of a second-order RC equivalent circuit model. The equivalent circuit model structure includes an ohmic resistor, an electrochemical polarization resistor, a deep polarization resistor, an electrochemical polarization capacitor, and a deep polarization capacitor.

[0129] The equivalent circuit model uses the least squares method to identify the parameters in the equivalent circuit model, and can collect the battery terminal current and terminal voltage data in real time during the simulation of the car driving process, and identify and correct the battery internal parameters online.

[0130] The observation equation is generated based on the Laplace transform of the battery's internal parameters and is expressed as:

[0131] Among them, U ocv For open circuit voltage, U out I is the terminal voltage, R0 is the ohmic resistance, C1 is the electrochemically polarized capacitor, R1 is the electrochemically polarized resistor, C2 is the deeply polarized capacitor, R2 is the deeply polarized resistor, and s is a plane in the Laplace transform process.

[0132] Step S720: Establish a lithium battery SOC prediction model to predict the optimal predicted SOC value at time t based on the observation equation;

[0133] The steps to establish a lithium battery SOC prediction model include the following:

[0134] An EKF prediction model is established, which estimates the SOC value and the predicted terminal voltage value at time t based on the SOC value of the previous time step, the initial SOC value, and the current measurement value.

[0135] The SABO algorithm is incorporated to correct measurement errors in battery terminal voltage and current.

[0136] Based on the measurement error correction, the prior estimate of the SOC value at time t, and the predicted terminal voltage value, the optimal SOC value at time t is output.

[0137] Step S730: Based on the optimal predicted SOC value, the optimal fuel consumption factor is calculated and generated by calling the hybrid vehicle energy management model to achieve the distribution of engine torque and motor torque; wherein, the calculation method of the optimal fuel consumption factor is as follows:

[0138] s(t)=s eq0 +ξ(SOC ref -SOC(t)); where s(t) is the real-time equivalent consumption factor, s eq0 For the equivalent fuel factor constant part, SOC ref Let ξ be the initial SOC value and ξ be the penalty factor.

[0139] Ultimately, s(t) serves as the real-time equivalent consumption factor and the optimal fuel consumption factor. It participates in the calculation and allocation of engine fuel consumption, motor equivalent fuel consumption, and the corresponding motor and engine power, thereby optimizing engine torque and motor torque and ultimately realizing the real-time application of energy management strategies.

[0140] This invention optimizes the energy management process of the Equivalent Minimum Fuel Strategy (ECMS). It simulates the dynamic changes of the lithium battery during driving using an equivalent circuit model, and collects and identifies internal battery parameters in real time. Building upon commonly used voltage and current parameters, it supplements this with real-time identification of parameters such as capacitance and resistance, providing more comprehensive source data for State of Charge (SOC) prediction and creating a data environment for real-time energy allocation. During SOC prediction, the EKF algorithm is optimized by incorporating the SABO algorithm to correct measurement errors, further improving the accuracy of the optimal SOC prediction value. This yields an accurate and optimal fuel consumption factor, enabling the best real-time application of the energy management strategy.

[0141] The above-disclosed embodiments are merely a few specific examples of the present invention. However, the present invention is not limited thereto, and any variations that can be conceived by those skilled in the art should fall within the protection scope of the present invention.

Claims

1. A fuel consumption distribution system based on SOC prediction, characterized in that, include: A hybrid electric vehicle energy management model is used to realize the constraint relationship between wheel torque, engine torque and motor torque. The constraint relationship refers to controlling the distribution ratio of engine torque and motor torque according to the optimal fuel consumption factor. An equivalent circuit model is used to simulate the dynamic changes of the lithium battery during driving. It collects the current and terminal voltage of the lithium battery during driving, acquires and outputs the battery's internal parameters, and establishes observation equations. The battery's internal parameters include terminal current, terminal voltage, open-circuit voltage, ohmic resistance, electrochemical polarization resistance, deep polarization resistance, electrochemical polarization capacitance, and deep polarization capacitance. The observation equations are generated based on the Laplace transform of the battery's internal parameters and are expressed as follows: ,in, For open circuit voltage, For current, For ohmic resistance, For electrochemically polarized capacitors, For electrochemical polarization resistance, For deep polarization capacitors, For deep polarization resistors, Let be a plane in the Laplace transform process; A lithium battery SOC prediction model is used to obtain the observation equation, and to calculate based on the observation equation. The optimal predicted SOC value at a given time; the lithium battery SOC prediction model includes: an EKF prediction model, used to estimate the optimal SOC value based on the previous SOC value, the initial SOC value, and the battery's internal parameters. The system includes: a predicted SOC value and terminal voltage at a given time; a SABO algorithm unit to correct measurement errors in battery terminal voltage and current; and a posterior estimation unit for performing posterior estimation based on the EKF algorithm, using the corrected current, terminal voltage, and prior estimates. The output shows the SOC value and predicted terminal voltage at each moment. The optimal SOC value at time t; The fuel consumption allocation module is used to calculate and generate the optimal fuel consumption factor based on the best predicted SOC value using the hybrid vehicle energy management model, thereby allocating engine torque and electric motor torque; wherein, the calculation method for the optimal fuel consumption factor is as follows: ;in, For real-time equivalent consumption factor, For the equivalent fuel factor constant part, For the initial SOC value, This is a penalty factor.

2. The fuel consumption distribution system according to claim 1, characterized in that, The equivalent circuit model is a vehicle battery model based on the Thevenin model with the addition of a second-order RC equivalent circuit model. The structural elements of the equivalent circuit model include ohmic resistors, electrochemical polarization resistors, deep polarization resistors, electrochemical polarization capacitors, and deep polarization capacitors.

3. The fuel consumption distribution system according to claim 2, characterized in that, The equivalent circuit model uses the least squares method to identify parameters, simulating the real-time acquisition of battery terminal current and voltage data during vehicle operation, and online identification and correction of specific values ​​of battery internal parameters.

4. A fuel consumption allocation method based on SOC prediction, characterized in that, Includes the following steps: A hybrid electric vehicle energy management model is established, in which the constraint relationship between wheel torque, engine torque and motor torque is realized. The constraint relationship refers to the control of the distribution ratio of engine torque and motor torque according to the optimal fuel consumption factor. An equivalent circuit model is established to simulate the dynamic changes of the lithium battery during driving. The current and terminal voltage of the lithium battery are collected during driving to obtain internal battery parameters, and an observation equation is established. These internal battery parameters include terminal current, terminal voltage, open-circuit voltage, ohmic resistance, electrochemical polarization resistance, deep polarization resistance, electrochemical polarization capacitance, and deep polarization capacitance. The observation equation is generated based on the Laplace transform of these internal battery parameters and is expressed as follows: ,in, For open circuit voltage, For current, For ohmic resistance, For electrochemically polarized capacitors, For electrochemical polarization resistance, For deep polarization capacitors, For deep polarization resistors, Let be a plane in the Laplace transform process; A lithium battery SOC prediction model is established to predict the SOC of the battery based on the observation equation. The optimal predicted SOC value at a given time; the establishment of the lithium battery SOC prediction model includes: establishing an EKF prediction model, which estimates the optimal SOC value based on the previous SOC value, the initial SOC value, and the current measurement value. The predicted SOC value and terminal voltage at each moment are calculated; the SABO algorithm is incorporated to correct measurement errors in battery terminal voltage and current; based on the measurement error correction and prior estimation... The output shows the SOC value and predicted terminal voltage at each moment. The optimal SOC value at time t; Based on the optimal predicted SOC value, the optimal fuel consumption factor is calculated using the hybrid vehicle energy management model to achieve the distribution of engine torque and electric motor torque; wherein, the calculation method of the optimal fuel consumption factor is as follows: ;in, For real-time equivalent consumption factor, For the equivalent fuel factor constant part, For the initial SOC value, This is a penalty factor.

5. The fuel consumption distribution method according to claim 4, characterized in that, The equivalent circuit model is a vehicle battery model based on the Thevenin model with the addition of a second-order RC equivalent circuit model. The equivalent circuit model structure includes an ohmic resistor, an electrochemical polarization resistor, a deep polarization resistor, an electrochemical polarization capacitor, and a deep polarization capacitor.

6. The fuel consumption allocation method according to claim 5, characterized in that, The least squares method is used to identify the parameters in the equivalent circuit model, and the battery terminal current and terminal voltage data can be collected in real time during the simulation of car driving, so as to identify and correct the internal parameters of the battery online.

Images