An automatic driving electric vehicle power system optimization method

Abstract

The application discloses an automatic driving electric vehicle power system optimization method, comprising the following steps: establishing a vehicle feedback system power model, which comprises a driver model, a monitoring controller model and a vehicle model; setting a driving cycle mode, and taking the speed limit, the parking position, the travel distance and the travel time in the driving cycle mode as the constraint conditions of the vehicle feedback system power model; simulating all power system schemes in the power system architecture through the whole vehicle model to obtain the optimal speed trajectory of each power system scheme, and the energy consumption and the peak power of the power system corresponding to each optimal speed trajectory; and selecting the power system scheme with the minimum energy consumption and the minimum peak power of the power system as the optimal power system scheme.

Description

Technical Field

[0001] This invention belongs to the field of electric vehicle powertrain technology, and specifically relates to an optimization method for the powertrain system of autonomous driving electric vehicles. Background Technology

[0002] Autonomous electric vehicles are considered the mainstream of future automobiles because they offer not only enhanced safety but also improved energy efficiency. This energy-saving optimization is achieved through improvements in the electric vehicle's powertrain. Electric vehicles utilize electrical energy storage, improving energy utilization efficiency and reducing energy waste by optimizing battery performance and reducing battery wear. Furthermore, electric vehicles can employ energy recovery technology during braking, converting braking energy into electrical energy stored in the battery, further improving energy utilization efficiency. In addition, optimized powertrains can achieve longer driving ranges. Optimizing motor power output and reducing battery self-discharge rates allow electric vehicles to achieve longer driving distances, further reducing energy waste and consumption during travel.

[0003] Currently, optimization methods for the powertrain systems of autonomous electric vehicles include: 1. Improving the energy density and charging efficiency of electric vehicles: This involves improving battery technology and battery management systems to increase battery energy density. However, increasing battery energy density requires the development of new battery and charging technologies, which is costly and requires long-term research and investment. Furthermore, increased battery energy density may also increase battery safety risks. 2. Reducing power loss by improving the motor, reducing friction in the transmission system, and lowering air resistance. However, this optimization method needs to be considered in the early stages of vehicle design and cannot be applied to existing electric vehicles, making it costly for existing models. 3. Employing a multi-motor drive system: This involves installing multiple motors in the electric vehicle and using intelligent control to adjust the operating status of each motor in real time. This method requires corresponding physical hardware and an intelligent control system, thus also presenting a cost problem. Summary of the Invention

[0004] The purpose of this invention is to provide a method for optimizing the power system of an autonomous electric vehicle. By establishing a dynamic model of the vehicle's feedback system, the optimal power system for the autonomous electric vehicle can be selected. The impact of various vehicle parameters on performance can be considered in the early design stage, without the need to improve the power system by using a higher-performance battery.

[0005] The technical solution provided by this invention is as follows:

[0006] A method for optimizing the powertrain system of an autonomous electric vehicle, comprising:

[0007] Establish a dynamic model of the vehicle feedback system, which includes: driver model, monitoring and controller model and vehicle model;

[0008] The driver model takes vehicle speed as input and outputs accelerator pedal opening and brake pedal opening.

[0009] The monitoring and controller model obtains the torque request signal of the traction motor and the brake line pressure signal based on the actual vehicle speed, accelerator pedal opening and brake pedal opening.

[0010] The vehicle model includes: powertrain architecture and subsystem component models;

[0011] The vehicle model takes into account the torque request signal of the traction motor and the brake line pressure signal, and outputs the vehicle's speed, position, and the status of the subsystem components.

[0012] A driving cycle mode is set, and the speed limit, parking position, travel distance, and travel time in the driving cycle mode are used as constraints on the vehicle feedback system dynamic model. All power system schemes in the power system architecture are simulated through the whole vehicle model to obtain the optimal speed trajectory of each power system scheme, and the energy consumption and peak power of the power system corresponding to each optimal speed trajectory are obtained. The power system scheme with the minimum energy consumption and the minimum peak power of the power system is selected as the optimal power system scheme.

[0013] Preferably, the driver model uses a PI controller; the difference between the current vehicle speed and the required vehicle speed is used as the input of the PI controller, and the output signals are the accelerator pedal opening and the brake pedal opening.

[0014] Preferably, the monitoring controller model includes a mode selection module and a mode operation module;

[0015] The mode selection generates a driving status command based on the accelerator pedal opening and the brake pedal opening. The mode operation module obtains the torque request signal of the traction motor and the brake line pressure signal based on the driving status command.

[0016] The driving states include: driving state and inertial coasting / braking / regenerative braking state.

[0017] Preferably, the powertrain architecture includes a front-wheel drive powertrain architecture and an all-wheel drive powertrain architecture.

[0018] Preferably, the subsystem component model includes: a motor model, a battery model, a gearbox model, an inverter model, a DC / DC converter model, and wheel and tire models.

[0019] Preferably, the type of motor in the motor model is an AC permanent magnet synchronous motor, and the equations simulating the motor are:

[0020]

[0021] Where, η em For motor efficiency, ω em T is the motor speed. em T is the torque required by the motor. load P is the output torque of the motor. em，elec The electric power of the motor is J, and ψ is the efficiency index. When consuming electricity, ψ = 1, and when charging, ψ = -1. em Let be the moment of inertia of the motor rotor. This represents the change in motor speed.

[0022] Preferably, in the battery model, based on the fast response of the battery voltage predicted by the current or temperature input, the battery is modeled as a zero-order equivalent circuit static model, and the battery is regarded as a constant voltage source coupled with internal resistance. The equation used to simulate the battery is:

[0023]

[0024] Among them, V out (t) represents the output voltage, V oc R0 is the open-circuit voltage, I is the battery internal resistance, and R0 is the open-circuit voltage. b (t) represents the battery current, SOC(t) represents the battery charge, SOC0 represents the initial battery charge, and C nom η is the nominal capacity of the battery. c (t) represents the Coulomb efficiency.

[0025] Preferably, in the gearbox model, the equations used to simulate the gearbox are:

[0026]

[0027] Among them, T out For the output torque, η t λ is the transmission efficiency, λ is the transmission ratio, and T is the transmission efficiency. in For the input torque, ω out For the output rotational speed, ω in For the input rotational speed, ω em，max V is the maximum speed of the motor. veh，max R is the vehicle's maximum speed. w This is the rolling radius of the wheel.

[0028] Preferably, the driving cycle modes include: UDDS city driving cycle mode and HWFET highway driving cycle mode.

[0029] Preferably, the optimal velocity trajectory is obtained based on two-dimensional dynamic programming of distance and velocity, including:

[0030] Let X be the two variables in the vehicle motion problem. k，i =[d k v i The initial and final conditions are X0 = [d0, v0] and X... f =[d f v f ];

[0031] Where d represents the vehicle distance, v represents the speed, i is the exponent of the speed, and k is the exponent of the distance;

[0032] The minimized cost function is:

[0033]

[0034] in, This is the minimum energy consumption stage solution for a vehicle traveling a unit distance in steps of 1. The vehicle accelerates at a step size 1 with a velocity of 1. arrive The power at that time, β is the compensation coefficient, Let the distance be d and the speed be k. The minimum energy consumption state solution at that time. The optimal acceleration input when the system is in its optimal state is the lead distance;

[0035] The value of β is calculated iteratively using the bisection method, with the vehicle model simulating states in steps of 1; the state update equation is:

[0036]

[0037] After calculating the optimal cost value, the simulation starts from the initial state of the vehicle to obtain the optimal speed trajectory.

[0038] The beneficial effects of this invention are:

[0039] The present invention provides an optimization method for the power system of autonomous electric vehicles, which establishes a vehicle feedback system power model including a driver model, a monitoring controller model, and a vehicle model; obtains the autonomous driving cycle through dynamic programming, and obtains the optimal power system for autonomous electric vehicles using the Pareto front method; through power system design and operation, battery life can be extended, charging efficiency can be improved, and energy consumption can be reduced, thereby increasing the vehicle's range and battery life, and reducing the vehicle's operating costs; the optimized power system scheme can improve the vehicle's dynamic performance and driving comfort, and enhance the passenger's travel experience. Attached Figure Description

[0040] Figure 1 This is a flowchart of the method for optimizing the powertrain system of an autonomous electric vehicle according to the present invention.

[0041] Figure 2 This is a schematic diagram of the driver model described in this invention.

[0042] Figure 3 This is a schematic diagram of the monitoring controller model described in this invention.

[0043] Figure 4 This is a schematic diagram of the mode selection module in the monitoring controller model described in this invention.

[0044] Figure 5 This is a schematic diagram of the mode operation module in the monitoring controller model described in this invention.

[0045] Figure 6 This is a schematic diagram of the power model of the automotive feedback system described in this invention.

[0046] Figure 7 This is a schematic diagram of the front-wheel drive vehicle model described in this invention.

[0047] Figure 8 This is a schematic diagram of the all-wheel drive vehicle model described in this invention. Detailed Implementation

[0048] The present invention will now be described in further detail with reference to the accompanying drawings, so that those skilled in the art can implement it based on the description.

[0049] like Figure 1 As shown, this invention provides a method for optimizing the powertrain system of an autonomous electric vehicle. The main implementation process is as follows: A vehicle feedback system dynamic model is established, which includes a driver model, a monitoring controller model, and a vehicle model. The vehicle feedback system dynamic model is developed using Simulink simulation software. Speed ​​limits, parking positions, travel distances, and travel times from the UDDS urban driving cycle and the HWFET highway driving cycle are used as constraints. A cycle is generated using a cost function and a state update equation. The optimal speed trajectory is obtained through simulation using the vehicle feedback system dynamic model. Feasible powertrain candidate schemes are screened based on performance requirements. The optimal powertrain scheme is selected using the Pareto front trade-off curve based on the objective function.

[0050] The driver model generates accelerator and brake pedal commands based on the current vehicle state and the desired vehicle state. For example... Figure 2As shown, in one embodiment, the driver model's driver program is modeled by a PI controller. The driver model's input signal is the reference vehicle speed minus the current vehicle speed, and its output signals are the accelerator pedal opening and the brake pedal opening.

[0051] The monitoring and controller model determines the operation of system components, and control commands depend on the driver's commands and the state of the vehicle and subsystem components. For example... Figure 3-5 As shown, the monitoring controller model is divided into two parts: mode selection and mode operation. In mode selection, based on the accelerator and brake pedal opening signals output by the driver and the actual vehicle speed provided by the vehicle model as input, the vehicle can be in two states: 1. Driving, 2. Coasting / Brake / Regenerative Braking. When parking, if the accelerator pedal opening is higher than a threshold, the vehicle enters the driving state. When the accelerator pedal opening is lower than a specific value or the brake pedal opening is higher than a specific value, the vehicle enters coasting, braking, or regenerative braking. In the coasting / braking / regenerative braking state, if the accelerator pedal opening is higher than a specific value and the brake pedal opening is lower than a specific value, the vehicle enters the driving state. During coasting or braking, if the brake pedal opening is higher than a specific value and the vehicle speed is lower than a specific value, the vehicle parks. The mode selection block issues accelerator or brake commands to the mode operation block based on the vehicle's state. The mode operation block converts these commands into traction motor torque requests and brake line pressure signals. When the vehicle is in motion, the mode selection block issues an accelerator command and simultaneously calculates the required torque based on the accelerator pedal opening output by the driver model and the accelerator-torque relationship diagram. When the vehicle is in a state of inertial coasting / braking / regenerative braking, the mode selection block issues a braking command and simultaneously calculates the required braking force based on the brake pedal opening output by the driver model and the braking-torque relationship diagram. The required braking force consists of the reverse required torque and the brake line pressure. The reverse required torque can charge the energy storage system, thus realizing the regenerative braking part.

[0052] The creation of the vehicle model includes selecting the powertrain architecture and constructing subsystem component models. The inputs to the vehicle model are the traction motor torque request signal and the brake line pressure signal from the controller, and the outputs are the vehicle's state, i.e., the vehicle's speed and position, as well as the state and individual variables of each component.

[0053] like Figure 6As shown, the vehicle's actual speed serves as the input signal to the driver model, and the driver model's output signals are the accelerator pedal opening and brake pedal opening. The accelerator and brake pedal openings serve as input signals to the monitoring controller model, and simultaneously, the vehicle's actual speed in the vehicle model also serves as an input signal to the monitoring controller model. In summary, the vehicle model's speed affects the driver model's pedal opening, and both the driver model's pedal opening and speed serve as input signals to the monitoring controller. The monitoring controller outputs torque commands to the motor and pressure commands to the brake lines to control vehicle operation. Furthermore, the vehicle speed and position can be used to derive the optimal speed trajectory for autonomous driving.

[0054] like Figure 7-8 As shown, this invention considers two architectures: front-wheel drive and all-wheel drive, and establishes vehicle models for front-wheel drive and all-wheel drive respectively.

[0055] The power system components that need to be modeled in the subsystem component model include: motor, energy storage system (battery), transmission system (gearbox), inverter, DC / DC converter, wheels and tires.

[0056] The methods for establishing the subsystem component models are as follows:

[0057] S1. The type of motor is an AC permanent magnet synchronous motor. The equation used to simulate the motor is:

[0058]

[0059] Where, η em For motor efficiency, ω em T is the motor speed. em For the required torque of the motor, T load P is the output torque of the motor. em，elec The electric power of the motor is J, and ψ is the efficiency index. When consuming electricity, ψ = 1, and when charging, ψ = -1. em Let be the moment of inertia of the motor rotor. This represents the change in motor speed. The data required for this model includes motor operating characteristics (peak and rated), efficiency / loss diagrams of the motor and generator, and motor data such as rotor inertia.

[0060] S2. Based on the fast response of battery voltage prediction using current or temperature input, the battery is modeled as a zero-order equivalent circuit static model. The modeling method treats the battery as a constant voltage source coupled with internal resistance. The equation used to simulate the battery is:

[0061]

[0062] Among them, V out (t) represents the output voltage, V ocR0 is the open-circuit voltage, I is the battery internal resistance, and R0 is the open-circuit voltage. b (t) represents the battery current, SOC(t) represents the battery charge, SOC0 represents the initial battery charge, and C nom η is the nominal capacity of the battery. c (t) represents the Coulomb efficiency.

[0063] When modeling the battery pack, all cells are considered to be identical, with the same battery capacity, and providing the same voltage and current. The equation used to simulate the battery pack is:

[0064]

[0065] Among them, V pack I is the battery pack voltage. pack For the battery pack current, C pack，nom n represents the nominal capacity of the battery pack. s n is the number of units connected in series. p This represents the number of parallel units.

[0066] S3. Based on the space, cost, and improved packaging of electric vehicles, the transmission and differential gears are combined and referred to as an electric drive gearbox. The gearbox is modeled as a simple gear reducer with constant efficiency and negligible inertia. The equations used to simulate the gearbox are:

[0067]

[0068] Among them, T out For the output torque, η t λ is the transmission efficiency, λ is the transmission ratio, and T is the transmission efficiency. in For the input torque, ω out For the output rotational speed, ω in For the input rotational speed, ω em，max V is the maximum speed of the motor. veh，max R is the vehicle's maximum speed. w This is the rolling radius of the wheel.

[0069] S4. Other components include the inverter, DC / DC converter, wheels, and tires. Since this model does not simulate thermal response, there is no cooling circuit model. The inverter controls the motor by converting DC power from the battery into AC power at the required frequency. The inverter is considered solely as an energy-consuming mechanism, with the same efficiency as the motor. The DC / DC converter converts the high-voltage source of the battery system to a low-voltage source to power the HVAC system, controllers, lighting fixtures, and charge the 12V lead-acid battery. The DC / DC converter is modeled as a constant energy-consuming mechanism between the ESS and auxiliary loads. The driveshaft is modeled as a rigid shaft. The wheels and tires are modeled with standard values ​​for passenger car tires for rolling radius and tire inertia, and rolling resistance coefficients for vehicles traveling on dry, well-maintained asphalt or concrete surfaces, ranging from 0.010 to 0.018. Other parameters are default values.

[0070] The input to the motor model is the required torque T. em and motor speed ω em The output is derived from the mode operation section of the monitoring controller model, where the motor speed is obtained based on the vehicle speed, wheel radius, and transmission ratio (wheel to rigid axle) in the monitoring controller model. The output is the motor output torque T. load and motor power P em，elec The method involves creating a Matlabfunction module in Simulink, inputting equations and unknowns into the module, and obtaining the output variables of the model.

[0071] The battery model was built using Simscape in Simulink. The battery voltage and charge output of the battery model are only used as monitoring parameters in the overall vehicle model, and the output variables are not used as output variables for other components. The battery's State of Charge (SOC) serves as a reference value for subsequent energy consumption.

[0072] The input to the transmission system (gearbox) model is the input torque T. in and input rotational speed ω in These variables come from the output of the motor connected to it, and the output is the output torque T. out and output speed ω out .

[0073] The wheels and brakes in other components are modeled using Simscape Tire blocks and Simscape Disk Brake blocks, respectively, while the motor, battery, and drive shaft connecting the wheels of the transmission system are modeled and connected using Simscape.

[0074] The longitudinal dynamics of the vehicle body were simulated using vehicle body blocks in Simscape. It is essentially a mask applied to the force balance equations of an aircraft. Key vehicle design parameters such as overall vehicle mass, drag coefficient, and frontal area are initialized here.

[0075]

[0076] This formula is for the vehicle body and requires the creation of a module. The input is the tire torque and radius to obtain Fw, and the output is Vveh. The constant value in the model is Meff, which represents the vehicle weight.

[0077] Faero=1 / 2ρCdA(Vveh) 2 Where ρ, Cd, and A are all constant values.

[0078] Froll = CrMg, where g is the acceleration due to gravity, and Cr is a constant.

[0079] Fg is ignored, and Faero, Froll, and Fg are also used as inputs to the model.

[0080] In the UDDS city driving cycle and HWFET highway driving cycle, the autonomous driving cycle period was derived separately, so the constraints on travel distance and time are the same as in the conventional cycle. The maximum speed depends on the speed limits along the route. The main focus is on acceleration constraints; to ensure comfort, acceleration is limited to 1.34 m / s². 2 The objective function's optimization criteria are fuel economy / energy consumption, powertrain cost, and weight; static constraints are primarily performance requirements for autonomous electric vehicles, including maximum speed, maximum cruising speed, acceleration performance, and hill-climbing ability. Two-dimensional dynamic programming based on distance and speed is used to calculate the optimal trajectory. The two variables for establishing the vehicle motion problem are X. k，i =[d k v i The vehicle distance is represented by d, and the speed by v. Speed ​​is a state variable, and distance is a stage variable. i is the speed exponent, representing the vehicle speed in the same stage under different cycle states, and k is the distance exponent, representing the distance traveled in the same stage of two cycles. The initial and final conditions are X0 = [d0, v0], X f =[d f v f The minimized cost function is:

[0081]

[0082] in, This is the minimum energy consumption stage solution for a vehicle traveling a unit distance in steps of 1. The vehicle accelerates at a step size 1 with a velocity of 1. arrive The power at that time, β is the compensation coefficient, This is the minimum energy consumption state solution when the distance d is k and the velocity v is i1. Let β be the optimal acceleration when the system is in its optimal state. The value of β is calculated iteratively using the bisection method. The vehicle model simulates states with a step size of 1, therefore the state update equation is:

[0083]

[0084] The above method evaluates the constraints of autonomous driving loops and discusses methods for generating loops. This method simulates full autonomy, where the vehicle has complete information about its future trajectory and constraints.

[0085] After calculating the optimal cost value, a simulation is performed starting from the vehicle's initial state to obtain the optimal speed trajectory. The loop generated after this dynamic programming process is called the autonomous driving drive loop. This trajectory is then used as input for optimizing the autonomous driving power system.

[0086] Using the Matlab Function module in Simulink, the existing velocity-displacement is used as the input signal. The corresponding Matlab code, namely the cost function and state update equation, is filled in to ensure that the cost value in the cost function is minimized. The Simulink model file is then compiled, and the velocity-displacement amount with the optimal velocity trajectory is finally output.

[0087] By screening feasible powertrain candidate schemes based on performance requirements, and using the autonomous driving drive cycle as input, the optimal speed trajectory of all feasible candidate schemes is plotted. A graph showing the relationship between energy consumption and peak power of the total powertrain system under all feasible candidate schemes is generated. To find the optimal candidate that minimizes the objective function, the peak power of the powertrain system is the independent variable, and the energy consumption of each feasible candidate scheme is the dependent variable in the graph. Considering that the powertrain system is more efficient at lower power levels in autonomous electric vehicles, the optimal power period for the front-wheel drive architecture is selected as 50-125 kW, and the optimal power period for the all-wheel drive architecture is selected as 75-200 kW, generating a Pareto front that meets the optimal power range and takes the lower energy consumption value. Based on the trade-offs between architecture, powertrain cost / power, and energy consumption, designers can select the powertrain system from the optimal powertrain system set with prominent Pareto fronts. This completes the powertrain optimization task for autonomous driving applications.

[0088] Example

[0089] Step 1: Analyze the motor characteristics based on the specific vehicle model data. Determine the powertrain architecture of the vehicle model initially based on the drive mode, number of motors, direct or indirect drive, number of gear reduction stages, and the position of the motor along the transmission system. This example considers both front-wheel drive (FWD) and all-wheel drive (AWD) architectures.

[0090] Step 2: Based on the specific vehicle model data, establish vehicle power system subsystem component models including motor, energy storage system, transmission system, inverter, DC / DC converter, wheels and tires, and establish corresponding equations.

[0091] Step 3: Use a PI controller to create a driver program for the driver model, such as... Figure 2 As shown, the specific model is as follows: The input signal for the driver model is the reference vehicle speed minus the vehicle speed, and the output signal is the accelerator and brake pedal opening. A monitoring controller model is established, as follows: Figure 3-5 As shown, control commands depend on the driver's commands and the state of the vehicle and subsystem components. The monitoring controller model is divided into two parts: mode selection and mode operation. In mode selection, based on the driver's commands and the vehicle's state, the vehicle can be in a stationary, propulsive, or coasting / regenerative braking mode. In mode operation, the command signals for the propulsion unit are different; the command signals are throttle and brake commands to the mode operation block, and the motor torque request can be positive or negative depending on the throttle and brake commands.

[0092] Step 4: Design the powertrain space, including the powertrain architecture and components. In this example, the architecture can be a front-wheel drive single-motor system or an all-wheel drive dual-motor system. Regarding components, six motors representing a wide power range were selected. The gear ratio for each combination was calculated based on the vehicle's maximum speed and the motors' maximum speeds. For the all-wheel drive system, the distinction between the front and rear wheels was ignored; for example, motor 3 connected to the front axle and motor 5 connected to the rear axle were assumed to be the same as motor 3 connected to the rear axle and motor 5 connected to the front axle. Since six motors were considered, there were six powertrain candidate schemes for FWD and 21 for AWD, for a total of 27 candidate schemes. The powertrain configurations for the candidate design space are shown in Table 1.

[0093] Table 1 Power System Configuration Table for Candidate Design Space

[0094] Candidate Solution # Front axle motor Rear axle motor Candidate Solution # Front axle motor Rear axle motor 1 1 - 15 4 2 2 2 - 16 5 2 3 3 - 17 6 2 4 4 - 18 3 3 5 5 - 19 4 3 6 6 - 20 5 3 7 1 1 21 6 3 8 2 1 22 4 4 9 3 1 23 5 4 10 4 1 24 6 4 11 5 1 25 5 5 12 6 1 26 6 5 13 2 2 27 6 6 14 3 2

[0095] Step 5: Consider the key performance requirements of autonomous electric vehicles, including maximum speed, maximum cruising speed, acceleration performance, and hill-climbing ability. Whether the powertrain can meet these requirements depends on vehicle characteristics such as mass, frontal area, drag coefficient, roll radius, and rolling resistance coefficient, as well as powertrain characteristics such as peak power and torque, rated power and torque, base / rated speed, and maximum speed. The performance requirements for autonomous driving are shown in Table 2.

[0096] Table 2 Performance Requirements for Autonomous Driving

[0097] Maximum speed [km / h] 135 Maximum cruising speed [km / h] 135 0-96 km / h acceleration time [s] 20 Climbing ability Maximum gradeability: 25%

[0098] Step 6: Derive the autonomous driving cycle based on the UDDS city driving cycle and the HWFET highway driving cycle. Assuming the vehicle has complete route and speed limit information, dynamic programming is used to calculate the optimal speed trajectory. Constraints obtained from the UDDS and HWFET cycles include speed limits, stopping positions, travel distance, and travel time. To ensure comfort, acceleration is limited to 1.34 m / s². 2 This ultimately yields the autonomous driving cycle.

[0099] Step 7: Filter feasible powertrain candidate schemes based on performance requirements. Using the autonomous driving cycle as input, generate the optimal speed trajectory for all feasible candidate schemes and plot the relationship between energy consumption and total powertrain peak power for all feasible candidate schemes. Since this example considers both FWD and AWD powertrain architectures, both curves need to be considered when generating the Pareto front. Finally, in the autonomous driving cycle scenario, plot the ratio of optimal energy consumption (architecture independent) to the energy consumption of each candidate scheme, expressed as a fraction. Based on the trade-offs between architecture, powertrain cost / power, and energy consumption, designers can select powertrains from the optimal powertrain set that stands out from the Pareto front. This completes the powertrain optimization task for autonomous driving applications.

[0100] Although embodiments of the present invention have been disclosed above, they are not limited to the applications listed in the specification and embodiments. They can be applied to various fields suitable for the present invention. For those skilled in the art, other modifications can be easily made. Therefore, without departing from the general concept defined by the claims and their equivalents, the present invention is not limited to the specific details and illustrations shown and described herein.

Claims

1. An automatic driving electric vehicle power system optimization method, characterized in that, The application relates to a method for optimizing a power system of a vehicle, comprising the following steps: establishing a power system dynamic model of the vehicle, which comprises a driver model, a supervisory controller model and a vehicle model; the driver model is input with a vehicle speed and output with a throttle pedal opening and a brake pedal opening; the supervisory controller model is input with the actual vehicle speed, the throttle pedal opening and the brake pedal opening, and output with a torque request signal of a traction motor and a brake pipe pressure signal; the vehicle model comprises a power system architecture and a subsystem component model; the vehicle model is input with the torque request signal of the traction motor and the brake pipe pressure signal, and output with a speed, a position and a state of a component of a subsystem of the vehicle; a driving cycle mode is set, and a speed limit, a parking position, a distance and a time in the driving cycle mode are taken as constraint conditions of the power system dynamic model of the vehicle; a whole vehicle model is simulated for all power system schemes in the power system architecture to obtain an optimal speed trajectory of each power system scheme, and to obtain an energy consumption and a peak power of the power system corresponding to each optimal speed trajectory; a power system scheme with the minimum energy consumption and the minimum peak power of the power system is selected as an optimal power system scheme; the optimal speed trajectory is obtained based on two-dimensional dynamic programming of distance and speed, and the two-dimensional dynamic programming comprises the following steps: Let X be the two variables in the vehicle motion problem. k,i =[d k ,v i The initial and final conditions are X0 = [d0, v0] and X... f =[d f ,v f ]; wherein d represents a vehicle distance, v represents a speed, i is an index of the speed, and k is an index of the distance; a minimized cost function is: wherein, is the minimum energy consumption phase solution for the vehicle to travel a unit distance with step size 1, is the power when the vehicle travels with step size 1 and acceleration a, and the speed is given by is the power when the vehicle travels with step size 1 and acceleration a, and the speed is given by is the minimum energy consumption state solution for the vehicle to travel a distance d with k steps, and the speed is given by is the minimum energy consumption state solution for the vehicle to travel a distance d with k steps, and the speed is given by is the optimal acceleration input when the system is in the optimal state solution is the look-ahead distance;​ a value of beta is obtained through bisection iterative calculation, and the vehicle model is simulated with a step length of 1; a state updating equation is: after the optimal cost value is calculated, the optimal speed trajectory is obtained by simulating from an initial state of the vehicle.

2. The method of claim 1, wherein, the driver model adopts a PI controller; a difference between a current vehicle speed and a required vehicle speed is taken as an input of the PI controller to obtain an output signal as the throttle pedal opening and the brake pedal opening.

3. The method of claim 2, wherein, the supervisory controller model comprises a mode selection module and a mode operation module; the mode selection generates a driving state command according to the throttle pedal opening and the brake pedal opening, and the mode operation module obtains the torque request signal of the traction motor and the brake pipe pressure signal according to the driving state command; wherein the driving state comprises a driving state and an inertia coasting / braking / regenerative braking state.

4. The method of claim 3, wherein, the power system architecture comprises a front-wheel drive power system architecture and a full-wheel drive power system architecture.

5. The method of claim 3 or 4, wherein, the subsystem component model comprises a motor model, a battery model, a gearbox model, an inverter model, a DC / DC converter model and a wheel and tire model.

6. The method of claim 5, wherein, the type of the motor in the motor model is an alternating current permanent magnet synchronous motor, and a motor simulation equation is: wherein η em is the motor efficiency, ω em is the motor rotational speed, T em is the motor required torque, T load is the motor output torque, P em,elec is the motor electric power, ψ is the efficiency index, ψ = 1 when discharging, and ψ = -1 when charging, J em is the moment of inertia of the motor rotor, is the change in the motor rotational speed.

7. The method of claim 6, wherein, in the battery model, a fast response of a battery voltage is predicted based on a current or temperature input, the battery is modeled as a zero-order equivalent circuit static model, the battery is regarded as a constant voltage source coupled with an internal resistance, and a battery simulation equation is: where V out (t) is the output voltage, V oc is the open circuit voltage, R0 is the internal resistance of the battery, I b (t) is the battery current, SOC(t) is the battery state of charge, SOC0 is the initial state of charge of the battery, C nom is the nominal capacity of the battery, η c (t) is the coulombic efficiency.

8. The method of claim 7, wherein, in the gearbox model, a gearbox simulation equation is: where T out is the output torque, η t is the transmission efficiency, λ is the transmission ratio, T in is the input torque, ω out is the output rotational speed, ω in is the input rotational speed, ω em,max is the maximum rotational speed of the motor, V veh,max is the maximum speed of the vehicle, R w is the rolling radius of the wheel.

9. The method of claim 8, wherein, the driving cycle mode comprises a UDDS urban driving cycle mode and a HWFET highway driving cycle mode.

Images