A hybrid electric vehicle predictive energy management method considering soc trajectory

Abstract

The application discloses a hybrid electric vehicle predictive energy management method considering SOC trajectory and belongs to the field of plug-in hybrid electric vehicle control. Firstly, a global target SOC trajectory is linearly solved based on a vehicle driving path and a target electric quantity consumption, and a dynamic programming algorithm is used to solve optimal torque distribution in a prediction time domain so as to obtain a reference SOC trajectory. Then, an optimal engine power is solved by using a linear quadratic output tracking controller with the SOC following effect and the optimal engine fuel consumption as targets. Finally, each power source demand torque is calculated based on rules, and the control quantity is sent to a corresponding controller to realize torque distribution. The application aims to solve the optimal torque distribution in the prediction time domain and realize the control of electric quantity consumption in the vehicle driving path, that is, the prediction information of future working conditions is fully utilized to realize the control of the target electric energy consumption and improve the vehicle economy.

Description

Technical Field

[0001] This invention belongs to the field of hybrid vehicle control, and specifically relates to a predictive energy management method for hybrid electric vehicles that takes into account the SOC trajectory. Background Technology

[0002] Plug-in hybrid electric vehicles (PHEVs) have garnered significant attention due to their long driving range and substantial energy-saving and emission-reduction potential. Energy management strategies play a crucial role in achieving these capabilities. Developing reasonable energy management methods is beneficial for further unlocking the energy-saving potential of PHEVs. For example, the invention patent published on January 10, 2023 (Publication No.: CN113479186B) describes an optimization method for energy management strategies in hybrid electric vehicles. This method uses a global optimization algorithm and machine learning algorithm to automatically calibrate control parameters, achieving automatic extraction of control parameters for hybrid electric vehicles. However, the real-time performance of the dynamic programming algorithm used in this invention cannot be guaranteed. Another example is the invention patent published on March 21, 2023 (Publication No.: CN115817452A), which describes a hybrid electric vehicle and its energy management method and device. This invention obtains the actual SOC of the power battery, determines the target SOC of the power battery, and then determines the operating mode of the hybrid electric vehicle and the engine operating mode, ultimately distributing the torque between the engine and the motor. However, this invention only considers the target value of SOC, without taking into account the SOC change trajectory during vehicle operation. Summary of the Invention

[0003] To address the shortcomings of existing technologies and further improve the energy-saving performance of plug-in hybrid electric vehicles, this invention proposes a predictive energy management method for hybrid electric vehicles that considers the State of Charge (SOC) trajectory, characterized by:

[0004] Step 1, SOC trajectory calculation: This step controls the energy consumption of the vehicle during its journey. Based on the vehicle's location information and path obtained from the cloud, and the battery SOC status obtained from the BMS system, combined with the vehicle's target energy consumption, a global target SOC trajectory is obtained through a linear solution method. This step is completed before the vehicle departs. Subsequently, based on the global target SOC trajectory and the predicted vehicle speed trajectory in the predicted time domain obtained from the cloud, a dynamic programming (DP) algorithm is introduced to solve for the most economically efficient torque distribution in the predicted time domain to obtain the reference SOC trajectory. This can be divided into Step 1(a) and Step 1(b).

[0005] Step 1(a): In a connected environment, the vehicle can obtain the total distance from the starting point to the destination. Based on the target power consumption transmitted by the cloud or the driver, the global target SOC trajectory can be obtained.

[0006] Step 1(b): Solve for the optimal power distribution scheme of the power system based on the vehicle speed trajectory in the prediction time domain, and use the corresponding battery SOC trajectory as the reference SOC trajectory; then use the DP algorithm to obtain the theoretical optimal torque distribution scheme and its SOC trajectory in the prediction time domain.

[0007] Step 2, Energy Management Strategy Based on Reference SOC Trajectory: With the reference SOC trajectory following effect and engine fuel consumption as objectives, a linear quadratic output tracker (LQR) is used to solve for the optimal engine power solution, thereby ensuring the real-time performance of the controller; then, based on the engine power demand, the desired engine operating point is selected on the engine optimal curve, the torque demand of each power source is solved, and the torque demand of the engine and motor is sent to the corresponding controllers ECU and MCU respectively to realize vehicle energy management.

[0008] Compared with the prior art, the present invention has the following beneficial effects:

[0009] 1. The predictive energy management method described in this patent can concentrate the engine operating point more in the low fuel consumption rate region of the engine while ensuring that the engine operating point is near the optimal curve, thereby improving engine efficiency while achieving SOC following.

[0010] 2. The predictive energy management method described in this patent can further improve engine efficiency and motor efficiency, thereby improving the overall system efficiency and reducing the vehicle's fuel consumption and electrical energy consumption. Attached Figure Description

[0011] The following description of the embodiments, taken in conjunction with the accompanying drawings, will make the embodiments readily understood, wherein:

[0012] Figure 1 This is a framework for a predictive energy management method for hybrid electric vehicles according to embodiments of the present invention;

[0013] Figure 2 The powertrain layout of a plug-in hybrid electric vehicle according to an embodiment of the present invention; Detailed Implementation

[0014] Embodiments of the present invention are described in detail below. Examples of these embodiments are shown in the accompanying drawings, wherein the same or similar reference numerals denote the same or similar elements or elements having the same or similar functions throughout. The embodiments described below with reference to the accompanying drawings are exemplary and are only used to explain the present invention, and should not be construed as limiting the present invention.

[0015] The following description, with reference to the accompanying drawings, outlines a framework for a predictive energy management method for hybrid electric vehicles that takes into account the SOC trajectory; however, the invention is not limited to these embodiments.

[0016] Reference Appendix Figure 1The predictive energy management method of this invention consists of two steps. Step one is to solve for the reference SOC trajectory, which includes two steps: Step one (a) is to plan the global target SOC based on network connectivity information and target power consumption; Step one (b) is to use the DP algorithm to solve for the optimal torque distribution in the prediction time domain based on the global target SOC trajectory to obtain the reference SOC trajectory; Step two is to realize predictive energy management based on the reference SOC trajectory, using the LQR algorithm to solve for the engine power demand and further realize the optimal torque distribution.

[0017] The predictive energy management method for hybrid electric vehicles that considers SOC trajectory described in this invention applies to the powertrain structure of plug-in hybrid electric vehicles (PHEVs). Figure 2 The engine is connected to the planetary carrier of the planetary gear set via a one-way clutch, and the planetary gear set achieves power splitting. The motor MG1 is connected to the sun gear of the planetary gear set, and the engine speed regulation function is achieved by the motor MG1. When the motor MG1 needs to be locked, the lock-up clutch can connect the motor MG1 to the frame. The motor MG2 is connected to the ring gear of the planetary gear set through a two-speed automatic transmission (AMT). Finally, the final drive transmits the power of the power system to the wheels, driving the vehicle.

[0018] Step 1, SOC trajectory calculation: This step controls the energy consumption of the vehicle during its journey. Based on the vehicle's location information, path, and battery SOC state obtained from the BMS system, and combined with the vehicle's target energy consumption, a global target SOC trajectory is obtained through a linear solution method. This step is completed before the vehicle departs. Subsequently, based on the global target SOC trajectory and the vehicle speed trajectory, a dynamic programming (DP) algorithm is introduced to find the most economical torque distribution to obtain the reference SOC trajectory. This can be divided into Step 1(a) and Step 1(b).

[0019] Step 1(a): In a connected environment, the vehicle can obtain the total distance from the starting position to the destination. Based on the target power consumption transmitted by the cloud or the driver, the global target SOC trajectory can be obtained according to Equation (1).

[0020]

[0021] In the formula, SOC all (d) represents the global target SOC corresponding to the vehicle's current position d; SOC initial The battery SOC at the vehicle's initial position; SOC end The expected battery SOC when the vehicle reaches its destination; d all This represents the total distance the vehicle travels from its starting point to its destination.

[0022] Step 1(b): Solve for the optimal power distribution scheme of the power system based on the vehicle speed trajectory, and use the corresponding battery SOC trajectory as the reference SOC trajectory; then use the DP algorithm to obtain the theoretical optimal torque distribution scheme and its SOC trajectory in the predicted time domain.

[0023] First, the prediction time domain is discretized into N. p The vehicle power requirement for each stage can be obtained through equation (2):

[0024]

[0025] In the formula, P req_k Let v be the total vehicle power requirement for stage k; eco_k To predict the vehicle speed in the k-th stage of the time domain; T tq The output torque of the power system; i0 is the transmission ratio of the main reducer; η T denoted as: η = mechanical efficiency of the transmission system; r = radius of the drive wheel; F0, F1, and F2 = coefficients of the constant, first-order, and second-order terms of the sliding resistance, respectively; v = vehicle speed; G = vehicle weight; i = road gradient; δ = vehicle rotational mass conversion factor; m = vehicle mass; dv = speed interval; dt = time interval.

[0026] The upper and lower boundaries of the battery SOC at each stage can be obtained by equation (3), which are the upper and lower limits corresponding to the current battery working at the maximum charging power or the maximum discharging power.

[0027]

[0028] In the formula, E bat R is the battery open-circuit voltage. bat P is the battery's internal resistance. bat Battery power;

[0029] To ensure the battery's state of charge (SOC) follows the global SOC and maintains vehicle economy, the state transition cost function is set as follows:

[0030]

[0031] In the formula, w dp_1 and w dp_2 These are the weighting coefficients for SOC following effect and vehicle economy, respectively.

[0032] The optimal solution in the prediction time domain can then be obtained as the control quantity that minimizes the sum of the cumulative cost functions at each stage:

[0033]

[0034] The constraints are:

[0035]

[0036] In the formula, ω eng_min and ω eng_max These are the engine's minimum and maximum speeds, respectively; T eng_min (ω eng_k ) and T eng_max (ω eng_k ) represent the minimum and maximum torques of the engine at the current speed, respectively; ω mg1_min and ω mg1_max These are the minimum and maximum speeds of motor MG1, respectively; T mg1_min (ω mg1_k ) and T mg1_max (ω mg1_k These represent the minimum and maximum torques of motor MG1 at the current speed, respectively.

[0037] Subsequently, the torque boundary range provided by the engine and motor MG1 across the entire speed range is discretized, and then constraints are added to the discrete points that do not meet the constraints using a penalty function:

[0038]

[0039] The state transition cost function at this point is as shown in equation (8):

[0040] L(x k ,u k )=w1f1(x k )+w2f2(u k )+w3f3(u k (8)

[0041] The constraints become:

[0042]

[0043] Based on this, the single-stage cost and global cumulative cost corresponding to different states and control quantities are calculated at each stage according to the set state transition cost function, and the forward solution of the entire stage is performed; then the reverse solution is performed to obtain the control sequence that optimizes the cumulative cost function, and the corresponding battery SOC trajectory is used as the reference SOC trajectory.

[0044] Step 2, Energy Management Strategy Based on Reference SOC Trajectory: With the reference SOC trajectory following effect and engine fuel consumption as objectives, a linear quadratic output tracker (LQR) is used to solve for the optimal engine power solution, thereby ensuring the real-time performance of the controller; then, based on the engine power demand, the desired engine operating point is selected on the engine optimal curve, the torque demand of each power source is solved, and the torque demand of the engine and motor is sent to the corresponding controllers ECU and MCU respectively to realize vehicle energy management;

[0045] In this step, the vehicle battery SOC is selected as the state variable; the battery power is the solution variable; and the battery SOC is selected as the output quantity. Therefore, the state-space expression of the system is:

[0046]

[0047] After discretizing and linearizing the state space, we perform a Taylor expansion at the system operating point P0 = 0, ignoring higher-order terms, and transform it into the standard form of a linear model.

[0048] To achieve better battery SOC tracking and reduce engine fuel consumption, the difference between battery SOC and reference SOC and engine fuel consumption rate are selected as performance indicators, and an objective function (11) is constructed:

[0049]

[0050] In the formula, SOC ref_k Here is the battery SOC reference value; con1 and con2 are the weighting coefficients corresponding to the battery SOC following effect and engine fuel consumption rate, respectively.

[0051] The expression between fuel consumption and power on the optimal curve of the engine obtained by fitting engine bench test data is shown in (12);

[0052]

[0053] In the formula, The engine fuel consumption at the optimal operating point of the engine curve; c1, c2, and c3 are the fitting coefficients; P eng Engine power;

[0054] Based on the relationship that the power required by the whole vehicle is provided by both the battery and the engine, the constant terms are combined, and the objective function of (11) above can be written as (13):

[0055]

[0056] Take u′ k Satisfying equation (14):

[0057]

[0058] Choosing a reference SOC as the desired output, the system error is:

[0059] e k =z k -y k =SOC ref_k -SOC k (15)

[0060] The objective function can be reduced to the standard form of the output tracking problem:

[0061]

[0062] The optimal control quantity u can be obtained by using a linear quadratic output tracking controller to solve the above objective problem. * The corresponding engine power requirement is:

[0063]

[0064] Find the operating point corresponding to the required power on the engine's optimal curve. At this point, the required torque of each power source can be calculated using equation ().

[0065]

[0066] The engine torque demand T is obtained. eng_req Motor MG1 requires torque T mg1_req MG2 motor requires torque T mg2_req The signals are sent to the ECU, MCU1, and MCU2 respectively to enable vehicle energy management.

[0067] This invention makes frequent use of terms such as DP and LQR, but the possibility of using other terms is not excluded. These terms are used merely for the convenience of describing and explaining the essence of this invention; interpreting them as any additional limitation would contradict the spirit of this invention.

[0068] Although embodiments of the invention have been shown and described, those skilled in the art will understand that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the claims and their equivalents.

Claims

1.A method for predicting energy management of a hybrid electric vehicle (HEV) considering state of charge (SOC) trajectory, comprising the following steps: Step 1: reference SOC trajectory solving: the control of the electric energy consumption during the driving process of the vehicle is realized, the global target SOC trajectory is obtained by a linear solving method based on the vehicle position information, the vehicle path and the battery SOC state obtained from the cloud, and the target electric energy consumption of the vehicle, and the calculation is completed before the vehicle departs; thereafter, the DP algorithm is introduced to solve the optimal torque distribution in the prediction time domain based on the global target SOC trajectory and the vehicle speed trajectory in the prediction time domain obtained from the cloud, to obtain the reference SOC trajectory, which can be divided into step 1(a) and step 1(b); Step 1(a): in a networked environment, the vehicle can obtain the total distance from the starting position to the destination, and based on the target electric energy consumption transmitted by the cloud or the driver, the global target SOC trajectory can be obtained; Step 1(b): the optimal power distribution scheme of the power system is solved based on the vehicle speed trajectory in the prediction time domain, and the corresponding battery SOC trajectory is taken as the reference SOC trajectory; thereafter, the DP algorithm is used to obtain the theoretically optimal torque distribution scheme in the prediction time domain and its SOC trajectory; Step 2: energy management strategy based on the reference SOC trajectory: the engine fuel consumption and the reference SOC trajectory following effect are taken as the target, the linear quadratic optimal output tracker (LQR) is used to solve the optimal solution of the engine power, and the real-time performance of the controller is ensured; thereafter, the expected engine operating point is selected on the engine optimal curve based on the engine demand power, the demand torque of each power source is solved, and the demand torque of the engine and the motor is sent to the corresponding controller ECU and MCU respectively, to realize the energy management of the vehicle. ​ ​ ​ ​ ​

Images