An event-triggered multi-objective MPC based energy management method for series hybrid electric vehicles

Abstract

The application discloses a series hybrid electric vehicle energy management method based on event-triggered multi-target MPC, and belongs to the technical field of hybrid electric vehicle energy management and motion control. Firstly, a power system and a dynamics system of the vehicle are modeled, and the obtained mathematical model is used as a prediction model of multi-target MPC. Then, according to a control target, a suitable optimization target function and system constraints are selected to construct a multi-target MPC constraint optimization problem. The system state at the current time is sampled, a finite time domain multi-target MPC constraint optimization problem is constructed and solved, and the optimal control sequence predicted at the current time is obtained, and the first s optimal controls are applied to the vehicle system. Finally, an event-triggered mechanism is designed, the new system state is obtained at a new triggering time, and the multi-target MPC constraint optimization problem is updated, and the rolling iteration is performed until the control process is finished. The application guarantees the stability of the power battery pack and the engine power of the series hybrid electric vehicle, and the calculation efficiency of the optimization problem is higher.

Description

Technical Field

[0001] This invention belongs to the field of hybrid electric vehicle energy management and motion control technology, specifically a series hybrid electric vehicle energy management method based on event-triggered multi-objective MPC. Background Technology

[0002] While the rapidly developing automotive industry has brought convenience to people's lives, it has also generated negative environmental and resource problems, such as deteriorating air quality and dwindling oil and gas resources. In recent years, energy conservation and emission reduction have become topics of international concern, and new energy vehicles, as an effective solution to the energy crisis, have experienced rapid growth in the automotive market. Electric vehicles, in particular, are gaining increasing market acceptance due to their numerous technological advantages, including environmental friendliness, economy, energy efficiency, and low noise. However, the high cost of electric vehicles remains a significant concern because high energy density and high power density energy storage technologies have not yet been effectively developed. Over the past decade, hybrid vehicles, combining the advantages of both traditional gasoline vehicles and electric vehicles, have become increasingly popular in the automotive market. Series hybrid vehicles, in particular, offer advantages such as high performance, low emissions, and flexible operating modes, significantly reducing fuel consumption.

[0003] For hybrid electric vehicles (HEVs), the powertrain typically includes two or more energy sources for propulsion, such as an internal combustion engine and a battery, making energy management more complex than in conventional vehicles. HEV energy management strategies can rationally allocate power from multiple power sources according to different operating conditions, reducing fuel consumption and achieving energy conservation and emission reduction. A sound energy management strategy is key to improving overall vehicle fuel economy; therefore, energy management strategies have received widespread attention and research. Considering energy consumption, economic costs, and weight, research on HEV energy management has significant engineering implications.

[0004] Energy management strategies for hybrid electric vehicles (HEVs) typically effectively allocate and optimize the power of the battery and engine, ensuring stable battery operation, meeting power demands under different operating conditions, and reducing fuel consumption. The main approach to energy management in HEVs is to rationally allocate power while satisfying state and input constraints. Currently, there are two main types of HEV energy management strategies:

[0005] 1) Rule-based energy management strategies, such as the rule-based multi-level energy allocation strategy proposed by João P. Trovão et al. and the fuzzy rule-based energy allocation strategy proposed by Chun-Yan L et al., but rule-based methods usually cannot guarantee optimal energy saving;

[0006] 2) Energy management strategies based on optimization. Stefano Di Cairano et al. used model predictive control to manage the energy distribution of series hybrid electric vehicles, using the battery to adjust the transition between different operating points to smooth engine transients and thus improve efficiency. Liangfei Xu et al. proposed a dynamic programming-based dual-loop Pareto optimal strategy to achieve minimum fuel economy and system durability.

[0007] The aforementioned optimization-based energy management strategies can only optimize energy consumption economy but cannot simultaneously optimize motion control, which greatly limits the application scope of energy management strategies. In particular, these strategies cannot guarantee energy allocation under different operating conditions. Furthermore, it can be found that most of these methods do not consider the actual constraints of system state and control input constraints, while the capacity of energy control input and motion control input is limited in practical applications. Optimization-based energy management strategies typically have high computational resource requirements, which undoubtedly brings new challenges; on the other hand, these control methods only satisfy certain performance indicators such as minimum energy consumption during the control process. Therefore, there is an urgent need for a method that can consider these factors, including energy consumption economy, motion control performance, vehicle motion constraints, and computational resources, while designing energy management strategies.

[0008] Event-triggered multi-objective model predictive control (MPC) methods can achieve a balance among multiple performance indices while satisfying constraints. It's worth noting that although Weida Wang, Yonggang Liu, and Liang Lia have studied using multi-objective optimization methods to solve energy management problems related to hybrid electric vehicles, these methods do not simultaneously consider the constraints of motion control performance indices and limited computational resources. Therefore, it is difficult to ensure a balance between system motion control, computational complexity, and energy management in practical applications. Furthermore, since different operating conditions have different requirements for power allocation, the introduction of an event-triggered prediction mechanism can clearly make predictive adjustments, management, and optimizations to power allocation, and further reduce the number of times the MPC optimization problem is solved, thus saving computational resources. Summary of the Invention

[0009] The purpose of this invention is to provide an energy management method for series hybrid electric vehicles based on event-triggered multi-objective MPC, which obtains optimal energy consumption economy and motion control while meeting constraints and power requirements, so as to solve the problems of actual constraints, limited computing resources, inability to meet given performance and high energy consumption in the energy management and control of series hybrid electric vehicles.

[0010] A series hybrid electric vehicle energy management method based on event-triggered multi-objective MPC includes the following steps:

[0011] Step 1: Model the powertrain and dynamics of the series hybrid electric vehicle to obtain the whole vehicle system model;

[0012] (1) The longitudinal dynamic model of the series hybrid electric vehicle is obtained by modeling the dynamic system, and is expressed as:

[0013]

[0014] in, For the overall quality of the vehicle, For the longitudinal speed of the vehicle; traction force of a hybrid vehicle Provided by a permanent magnet synchronous motor; For rolling resistance, The rolling resistance coefficient, The road surface inclination angle; For air resistance, For windward area, This refers to the air drag coefficient; For climbing resistance; It is the gravitational constant;

[0015] (2) The energy flow model of the series hybrid electric vehicle power system is as follows:

[0016] The energy flow of the power system includes: the energy from the combustion of fuel in the internal combustion engine is provided to the engine, which in turn provides energy to the permanent magnet synchronous generator; the electrical energy output from the permanent magnet synchronous generator and the power battery pack is input into the electric motor after passing through a converter, or the energy output from the permanent magnet synchronous generator is input into the power battery pack for storage after passing through a converter; the electric motor transmits energy to the transmission system, which drives the series hybrid electric vehicle.

[0017] The overall energy flow model of the dynamic system is as follows:

[0018]

[0019] in, The transmission power of the transmission system. Let be the energy conversion efficiency constant of the converter. Let be the efficiency constant of the electric motor. This represents the constant transmission efficiency constant of the transmission system. For symbolic functions, For the power battery pack power, This refers to the power output of a permanent magnet synchronous generator. This represents the charge / discharge efficiency constant of the power battery pack.

[0020] (3) Establish a vehicle system model based on the dynamics model and the power model;

[0021] Permanent magnet synchronous generator power Power of the battery pack The traction force of the permanent magnet synchronous generator is The traction force of the power battery pack is .

[0022] Select the vehicle system status as The control input is Then, the complete vehicle system model, including the motion control and powertrain models, is obtained as follows:

[0023]

[0024] in, , For traction power, This is due to braking energy loss. Battery pack state of charge. Indicates the remaining battery power. This is the battery open-circuit voltage. For battery pack resistance, This refers to the battery's rated capacity. This refers to the fuel flow rate of an internal combustion engine. This represents the initial fuel flow rate of the internal combustion engine under no-load conditions. As gasoline has a low calorific value, It is the energy conversion constant. This indicates the vehicle's location.

[0025] The above-mentioned vehicle system can be represented by a general nonlinear function as follows:

[0026]

[0027] in, .

[0028] Step 2: Construct the objective function for the multi-objective MPC-constrained optimization problem based on the vehicle system model;

[0029] current The objective function of the time-bounded multi-objective MPC-constrained optimization problem is:

[0030]

[0031] in, To predict the time domain, For the system in Always Prediction of time error state, For the system in Always Predictions that control input at all times. For the phased objective function

[0032]

[0033] in, These are the weighting coefficients. For the target speed, For the target location, This is a reference value for the battery status.

[0034] Step 3: Set the constraints for the multi-objective MPC constraint optimization problem, and solve the problem based on the model's prediction of the system's future dynamics. The multi-objective MPC-constrained optimization problem at time 1000 is obtained. The optimal control sequence predicted at each time step;

[0035] The constraints include:

[0036] 1) Initial state, .

[0037] 2) Speed ​​constraints, ,in, This is the vehicle's maximum speed.

[0038] 3) Traction constraint, ,in, This is the maximum traction force.

[0039] 4) Battery state constraints. ,in, , These represent the minimum and maximum battery capacities, respectively.

[0040] 5) Generator power constraints, ,in, This is the generator's maximum power. This is the minimum power.

[0041] 6) Battery power constraint, , in, This is the battery's maximum power. This is the minimum power.

[0042] 7) Control input constraints, , .

[0043] Setting the prediction time domain for multi-objective MPC and sampling interval The weight parameters of the objective function and other relevant parameters. Set the reference position. and trajectory velocity The initial state of the hybrid vehicle is set as follows: Based on the above constraints, the solution is obtained. The specific form of the time-limited optimization problem is:

[0044]

[0045]

[0046]

[0047]

[0048]

[0049]

[0050]

[0051]

[0052]

[0053]

[0054] Step four, place the front An optimal control sequence is applied to the vehicle system, and an event triggering mechanism is designed to determine the next triggering time. ;

[0055] The event triggering mechanism is as follows:

[0056]

[0057] in, For the next triggering time, In order to be in The actual state after acting on the whole vehicle system, constant This is the trigger threshold. , The optimal state and control input are obtained from the prediction.

[0058] Step 5, when the new trigger moment When the new sampled state arrives, the new sampled state is used as the initial state to construct a new multi-objective MPC constrained optimization problem at the new moment. Then, return to step three and iterate continuously until the control process ends.

[0059] The beneficial effects of this invention are mainly reflected in:

[0060] (1) This invention utilizes an event-triggered multi-objective MPC method to optimize motion control and energy management of a series hybrid electric vehicle. The designed objective function simultaneously considers the vehicle's motion control objective, battery energy consumption, and fuel energy consumption. By optimizing this objective, power distribution between the power battery pack and the engine can be achieved while satisfying motion control requirements. Furthermore, specific performance indicators can be further optimized by setting weight parameters.

[0061] (2) The event-triggered multi-objective MPC motion control and energy management strategy designed in this invention takes the actual input range and state range as hard constraints of the optimization problem, which is more in line with actual applications; especially under different working conditions, such as start-stop, uphill, and acceleration, the power battery pack and engine can provide stable power.

[0062] (3) The event-triggered multi-objective MPC method designed in this invention can obtain the optimal control input in a shorter prediction time domain while satisfying the constraints, and reduce the number of times to solve the optimization problem, greatly saving the amount of computation, so that the proposed method can be applied in real time on hybrid electric vehicles. Attached Figure Description

[0063] Figure 1 This is a schematic diagram of the energy flow and motion control framework for a series hybrid electric vehicle according to the present invention.

[0064] Figure 2 This is a flowchart of the energy management method for series hybrid electric vehicles based on event-triggered multi-objective MPC according to the present invention. Detailed Implementation

[0065] The present invention will now be described in detail with reference to the accompanying drawings and specific embodiments. The embodiments described are exemplary and intended to explain the present invention, but should not be construed as limiting the present invention.

[0066] This invention proposes an energy management method for series hybrid electric vehicles based on event-triggered multi-objective MPC, such as... Figure 2 As shown, the principle involves modeling the vehicle's powertrain and dynamics systems, primarily including the battery pack, engine, generator, and vehicle dynamics model. The resulting mathematical model is then used as a prediction model for multi-objective MPC (Multi-Process Control), and a suitable objective function and system constraints are selected based on the control objective to construct a multi-objective MPC constrained optimization problem. The system state at the current moment is sampled to construct a finite-time domain multi-objective MPC constrained optimization problem; an event-triggered mechanism is designed; the constrained optimization problem is solved to obtain the predicted optimal control sequence at the current moment, and the next trigger moment is determined based on the event-triggered mechanism. , will go An optimal control is applied to the vehicle system; at each new trigger moment, a new system state is acquired and updated. The multi-objective MPC constraint optimization problem is iterated over and over until the control process ends.

[0067] like Figure 1 As shown, the implementation of the present invention is mainly divided into two parts. The first part is to model the power system and dynamics system of the vehicle, mainly including the power battery pack, engine, generator and vehicle dynamics model, etc. The second part is to use the mathematical model obtained in the first part as the prediction model of multi-objective MPC, and then select appropriate optimization objective function and system constraints according to the control objective to construct a multi-objective MPC constrained optimization problem. In addition, an event triggering mechanism is constructed to solve the MPC optimization problem once and apply multiple optimal control inputs to the vehicle system.

[0068] Taking a series hybrid electric vehicle as an example, the specific steps of the method of the present invention are as follows:

[0069] Step 1: Modeling the vehicle's powertrain and dynamics systems and selecting the corresponding parameters;

[0070] Step 1.1: The longitudinal dynamics model of a series hybrid electric vehicle can be expressed as:

[0071]

[0072] in, For the overall quality of the vehicle, Let be the longitudinal speed of the vehicle. For rolling resistance, The rolling resistance coefficient, The road surface inclination angle; For air resistance, For windward area, The air drag coefficient, For hill-climbing resistance, the traction of hybrid electric vehicles is provided by a permanent magnet synchronous motor.

[0073]

[0074] in, For traction power, Indicates braking energy loss. This refers to the transmission power of the transmission system.

[0075] Step 1.2: Determine the power balance equation between the transmission power and the motor power of the transmission system as follows:

[0076]

[0077] in, For the mechanical power of the electric motor, For a constant transmission efficiency constant, It is a symbolic function.

[0078] Step 1.3: Determine the power balance relationship between the DC converter and the motor, i.e.

[0079]

[0080] in, For converter power, is the efficiency constant of the electric motor.

[0081] Step 1.4: Determine the power balance relationship between the converter, permanent magnet synchronous generator, and power battery pack:

[0082]

[0083] in, Let be the energy conversion efficiency constant of the converter. For the power battery pack power, This refers to the power of a permanent magnet synchronous generator.

[0084] Step 1.5: The energy modeling between the engine and the permanent magnet synchronous generator is as follows:

[0085]

[0086] in, is the efficiency constant of the permanent magnet synchronous generator. The power provided to the engine.

[0087] Step 1.6: The relationship between the internal combustion engine fuel flow rate and the permanent magnet synchronous generator can be modeled as follows:

[0088]

[0089] in, This represents the initial fuel flow rate of the internal combustion engine under no-load conditions. As gasoline has a low calorific value, is the energy conversion constant.

[0090] Step 1.7: Determine the state estimation model for the power battery pack as follows:

[0091]

[0092] Where: the state of charge of the battery pack Indicates the remaining battery power. For the battery's rated capacity, This is the charging and discharging current; during charging, During discharge, , This is the charge / discharge efficiency constant. It is derived from Kirchhoff's voltage law.

[0093]

[0094] and battery output power

[0095]

[0096] The estimated state of the battery pack can be obtained.

[0097]

[0098] in, This is the battery open-circuit voltage. For battery pack resistance, This is the closed-circuit voltage of the battery. .

[0099] Step 1.8: Based on steps 1.1–1.7, the overall energy flow model can be obtained.

[0100] .

[0101] Step 1.9: The power of the permanent magnet synchronous generator and the battery power can be respectively determined by... and The calculation yielded, where , Select the vehicle system status as The control input is This allows us to obtain a complete vehicle system model that includes motion control and powertrain models.

[0102]

[0103] in, The aforementioned vehicle system can be represented by a general nonlinear function as follows:

[0104]

[0105] in, .

[0106] Step 2: Implementation of motion control and energy management strategies based on multi-objective MPC;

[0107] Step 2.1: Set the current The objective function of the time-bounded multi-objective MPC-constrained optimization problem is:

[0108]

[0109] in, To predict the time domain, For the system in Always Prediction of time error state, For the system in Always Predictions that control input at all times. For the phased objective function

[0110]

[0111] in, These are the weighting coefficients. For the target speed, For the target location, This is a reference value for the battery status.

[0112] Step 2.2: Set constraints for the multi-objective MPC optimization problem.

[0113] Initial state, .

[0114] Speed ​​constraints ,in, This is the vehicle's maximum speed.

[0115] Traction constraint, , in, This is the maximum traction force.

[0116] Battery state constraints, ,in, , These represent the minimum and maximum battery capacities, respectively.

[0117] Generator power constraints , in, This represents the generator's maximum power. This is the minimum power.

[0118] Battery power constraints , in, This is the battery's maximum power. This is the minimum power.

[0119] Control input constraints, , .

[0120] Step 2.3: Set the prediction time domain for multi-objective MPC Sampling interval The objective function's weight parameters and other relevant parameters.

[0121] Step 2.4: Set the reference position trajectory velocity The initial state of the hybrid vehicle is set as follows: .

[0122] Step 2.5: Based on the model's prediction of the system's future dynamics, solve the problem. The multi-objective MPC-constrained optimization problem at time 1000 is obtained. The optimal control sequence predicted at each time step. The specific form of the time-limited optimization problem is:

[0123]

[0124]

[0125]

[0126]

[0127]

[0128]

[0129]

[0130]

[0131]

[0132]

[0133] Step 2.6: Construct the event triggering mechanism as follows

[0134]

[0135] in, For the next triggering time, In order to be in The actual state of the system after the action, constant This is the trigger threshold.

[0136] Step 2.7: Calculate the previous... Optimal control It acts on the system until a new triggering time. When the new sampled state arrives, the new sampled state is used as the initial state to construct a new multi-objective MPC constrained optimization problem at the new moment. Then, return to step 2.5 and iterate continuously until the control process ends.

Claims

1. A series hybrid electric vehicle energy management method based on event-triggered multi-objective MPC, characterized in that, Includes the following steps: Step 1: Model the powertrain and dynamics of the series hybrid electric vehicle to obtain the whole vehicle system model; Permanent magnet synchronous generator power , power battery pack power where , ; select the vehicle system state as , the control input is , the vehicle system model including the motion control and power system model is obtained: in, , For traction power, For braking energy loss, Transmission power of the drive system; state of charge of the battery pack Indicates the remaining battery power. This is the battery open-circuit voltage. For battery pack resistance, This refers to the battery's rated capacity. This refers to the fuel flow rate of an internal combustion engine. This represents the initial fuel flow rate of the internal combustion engine under no-load conditions. Because gasoline has a low calorific value, It is the energy conversion constant. For vehicle location; For the overall quality of the vehicle, The longitudinal speed of the vehicle; For rolling resistance, The rolling resistance coefficient, The road surface inclination angle; For air resistance, For windward area, This refers to the air drag coefficient; For climbing resistance; For the traction force of the permanent magnet synchronous generator, For the traction force of the power battery pack; The above-mentioned vehicle system can be represented by a general nonlinear function as follows: wherein ; Step 2: Construct the objective function for the multi-objective MPC-constrained optimization problem based on the vehicle system model; Current The objective function of the multi-objective MPC constrained optimization problem at the current time instant is: wherein, is a prediction of the time horizon, is a prediction of the time horizon, is a prediction of the time horizon, is a prediction of the time horizon, is a prediction of the time horizon, is a prediction of the time horizon, is a prediction of the time horizon; is a stage-wise objective function wherein, is a weight coefficient, is a target speed, is a target position, is a battery state reference value; Step three, set the constraint conditions of the multi-objective MPC constrained optimization problem, solve the optimal control sequence of the future dynamics of the model predictive system according to the multi-objective MPC constrained optimization problem at the time t ; Step four, set the constraint conditions of the multi-objective MPC constrained optimization problem, solve the optimal control sequence of the future dynamics of the model predictive system according to the multi-objective MPC constrained optimization problem at the time t ; Setting the prediction horizon of multi-objective MPC and sampling interval , weight parameters of objective functions and other corresponding parameters; setting reference position and trajectory velocity , setting initial state of hybrid electric vehicle as , and solving to obtain the specific form of optimization problem at time wherein, is the traction force of the hybrid vehicle; The constraints include: 1) Initial state, ; 2) speed constraints, wherein, Vmax is the vehicle maximum speed; 3) a tractive effort constraint, wherein, is the maximum tractive effort; 4) battery state constraints, wherein, , are the minimum and maximum battery capacity values, respectively. 5) generator power constraint, wherein, Pmaxis the maximum power of the generator, Pminis the minimum power; 6) battery power constraints, , wherein, is the maximum power of the battery, is the minimum power; 7) control input constraints, ;​ Step four, place the front An optimal control sequence is applied to the vehicle system, and an event triggering mechanism is designed to determine the next triggering time. ; The event triggering mechanism is as follows: wherein, is the next triggering time, is the triggering threshold; is the actual state after the action on the whole vehicle system, constant is the triggering threshold; , is the predicted optimal state and control input; Step five, when a new trigger time comes, build a new time's multi-objective MPC constrained optimization problem with the new sampled state as the initial state, go back to step three, and roll the iteration until the end of the control process.

2. The event-triggered multi-objective MPC based energy management method for series hybrid electric vehicle according to claim 1, wherein, The dynamic system modeling of a series hybrid electric vehicle yields a longitudinal dynamic model, which is expressed as: wherein, is the mass of the vehicle, is the longitudinal speed of the vehicle; is the rolling resistance, is the rolling resistance coefficient, is the road inclination; is the air resistance, is the frontal area, is the air resistance coefficient; is the climbing resistance; traction of the hybrid vehicle is provided by the permanent magnet synchronous motor.

3. The event-triggered multi-objective MPC based energy management method for series hybrid electric vehicle according to claim 1, wherein, The energy flow model for a series hybrid electric vehicle powertrain is as follows: wherein, is the transmission power of the transmission system, is the energy conversion efficiency constant of the converter, is the electric motor efficiency constant, is the constant transmission efficiency constant of the transmission system, is the sign function, is the power battery pack power, is the permanent magnet synchronous generator power, is the power battery pack charge and discharge efficiency constant.

4. The event-triggered multi-objective MPC based energy management method for series hybrid electric vehicle according to claim 3, wherein, The energy flow of the power system includes: the energy from the combustion of fuel in the internal combustion engine is provided to the engine, which in turn provides energy to the permanent magnet synchronous generator; the electrical energy output from the permanent magnet synchronous generator and the power battery pack is input to the electric motor after passing through a converter, or the energy output from the permanent magnet synchronous generator is input to the power battery pack after passing through a converter; the electric motor transmits energy to the transmission system, which drives the series hybrid electric vehicle.

Images