Energy management method, device, apparatus and storage medium

Abstract

The application discloses an energy management method, device, equipment and storage medium, and belongs to the field of vehicle engineering. The method comprises the following steps: identifying a current working condition of a vehicle according to multi-source sensing information of the vehicle; adjusting a target function of energy management and adjusting a constraint condition of vehicle control according to the current working condition, the target function being used for balancing economy, battery temperature and driving safety of the energy management; and generating a control instruction according to the target function and the constraint condition, the control instruction being used for realizing energy management meeting the target function and the constraint condition. The method can dynamically adjust an energy distribution strategy according to a vehicle working condition.

Description

Technical Field

[0001] This application relates to the field of vehicle engineering, and in particular to an energy management method, apparatus, device, and storage medium. Background Technology

[0002] With the rapid development and widespread adoption of the new energy vehicle industry, vehicle energy efficiency has become a core indicator for measuring the overall performance of a vehicle. Energy Management Strategy (EMS), as the "brain" of the new energy vehicle power system, directly determines the vehicle's economy and power performance.

[0003] The energy management strategies adopted by new energy vehicles in related technologies rely on engineering experience to formulate fixed or semi-fixed control rules, such as calibration and parameter setting based on standard cycle conditions such as NEDC (New European Driving Cycle) and WLTC (Worldwide Light-duty Test Cycle).

[0004] However, the standardized operating conditions differ significantly from extreme natural environments such as extreme cold, high temperature, and high altitude in actual driving, as well as from the drastic changes in urban traffic congestion with dense traffic flow and frequent starts and stops. This makes it difficult for energy management strategies to adapt to non-standard operating conditions, and the rationality of energy regulation is greatly reduced. Summary of the Invention

[0005] This application provides an energy management method, apparatus, device, and storage medium, which can dynamically adjust the energy distribution strategy according to vehicle operating conditions. The technical solution is as follows: According to one aspect of this application, an energy management method is provided, the method comprising: Based on the vehicle's multi-source perception information, the current operating condition of the vehicle is identified; The objective function of energy management and the constraints of vehicle control are adjusted according to the current operating conditions. The objective function is used to balance the economy of energy management, battery temperature and driving safety. Control commands are generated based on the objective function and the constraints, and these control commands are used to implement energy management that conforms to the objective function and the constraints.

[0006] Optionally, the objective function includes a weighted sum of energy consumption cost indicators, battery temperature maintenance indicators, and driving safety indicators; The objective function for adjusting energy management based on the current operating conditions includes at least one of the following: Adjust the energy consumption cost index corresponding to the energy consumption economic weight according to the current operating conditions; Adjust the battery temperature maintenance weight corresponding to the battery temperature maintenance index according to the current operating conditions. The driving safety weights of the driving safety indicators are adjusted according to the current operating conditions.

[0007] Optionally, the initial value of the economic efficiency weight is a first value; the initial value of the battery temperature maintenance weight is a second value; and the initial value of the driving safety weight is a third value, wherein the first value is greater than the second value and the first value is greater than the third value. The objective function for adjusting energy management based on the current operating conditions includes at least one of the following: When the current working condition is a complex working condition with low temperature and low adhesion, the energy consumption cost index corresponding to the energy consumption economy weight is reduced from the first value to the fourth value. When the current working condition is a complex working condition of low temperature and low adhesion, the weight of the battery temperature maintenance index is increased from the second value to the fifth value. When the current working condition is a complex working condition with low temperature and low adhesion, the driving safety weight corresponding to the driving safety index is increased from the third value to the sixth value. The fifth value is greater than the sixth value, and the sixth value is greater than the fourth value.

[0008] Optionally, adjusting the constraints for vehicle control based on the current operating conditions includes: Adjust the thermal management strategy according to the current operating conditions; and / or, Adjust torque distribution and anti-slip control according to the current operating conditions.

[0009] Optionally, the constraints for adjusting vehicle control include: When the current operating condition is a complex condition of low temperature and low adhesion, at least one of the following thermal management strategies shall be implemented: Shut down the battery cooling circuit; Open the coolant series valve between the motor controller MCU and the battery pack; When the motor is not working, the positive temperature coefficient PTC heater is instructed to heat the coolant at maximum power. The target curve for battery temperature rise is adjusted to raise the battery temperature to a first temperature within a first duration; the first temperature is determined based on the limiting temperature of the battery charging power.

[0010] Optionally, adjusting torque distribution and anti-slip control according to the current operating conditions includes: Under the current operating condition of low temperature and low adhesion complex conditions, a first-order low-pass filter is applied to the requested torque corresponding to the accelerator pedal request; and the system response time coefficient of the accelerator pedal is increased: the time coefficient is used to control the speed at which the system responds to the accelerator pedal request; and / or, Under the current operating condition of low temperature and low adhesion, the maximum adhesion of the wheels is dynamically calculated based on the real-time estimated road adhesion coefficient; the output torque of the motor is limited based on the maximum adhesion and a safety factor; and / or, When the current working condition is a complex condition of low temperature and low adhesion, and the slip ratio of the first wheel is detected to be higher than the slip ratio threshold, the output torque of the first motor is reduced and the output torque of the second electrode is increased.

[0011] Optionally, generating control instructions based on the objective function and the constraints includes: The objective function and the constraints are input into the model predictive controller (MPC), which generates the control commands based on the vehicle's current state and operation requests. The predictive controller generates a control sequence in the rolling time domain, and the control sequence includes at least one control command.

[0012] Optionally, the energy consumption cost index of the objective function includes a vehicle speed tracking deviation term; the battery temperature maintenance index of the objective function includes a thermal coupling cost term; the driving safety index of the objective function includes a desired acceleration deviation term and a motor torque increment penalty term; the objective function is used to evaluate the quality of control commands; the vehicle speed tracking deviation term is used to penalize the deviation between the actual vehicle speed and the target vehicle speed; the desired acceleration deviation term is used to penalize the deviation between the actual acceleration and the desired acceleration; the motor torque increment penalty term is used to penalize the change in motor torque within adjacent control cycles; the thermal coupling cost term is used to penalize the deviation between the actual power and the target power of the PTC heater. The step of inputting the objective function and the constraints into the model predictive controller (MPC) to generate the control instructions includes: Based on the constraints and with the objective function as the goal, the online quadratic programming solver of the MPC iteratively solves the problem in the rolling time domain based on the current state of the vehicle and the operation request, and outputs a control sequence.

[0013] Optionally, the control commands include at least one of the following: target motor torque, PTC heating power, and ESP intervention threshold.

[0014] Optionally, the method further includes: The vehicle's center of gravity speed is collected by the chassis domain controller at the first sampling frequency, and the longitudinal slip ratio is calculated. When the longitudinal slip ratio is in the subcritical range, the objective function and / or the constraints in the MPC are softly modified.

[0015] Optionally, the soft modification of the objective function and / or constraints in the MPC includes at least one of the following: Reduce the torque increment step size; Adjust the weight of at least one of the following in the objective function: vehicle speed tracking deviation, thermal coupling cost, expected acceleration deviation, and motor torque increment penalty. Adjust the boundaries of the constraints.

[0016] Optionally, the weights corresponding to at least one of the vehicle speed tracking deviation term, thermal coupling cost term, desired acceleration deviation term, and motor torque increment penalty term in the adjustment objective function include: Increase the weight of the expected acceleration deviation term and decrease the weight of the vehicle speed tracking deviation term.

[0017] Optionally, adjusting the boundaries of the constraints includes: When the longitudinal slip ratio is increasing, the maximum friction force of the virtual ground is reduced in advance; the maximum friction force of the virtual ground is used to constrain the maximum value of the longitudinal force of the wheel.

[0018] Optionally, the method further includes: The vehicle's center of gravity speed is collected by the chassis domain controller at the first sampling frequency, and the longitudinal slip ratio is calculated. If the longitudinal slip ratio is higher than the slip ratio threshold, the control command output by the MPC is not executed, and an emergency intervention command is executed through the hardware safety monitoring unit (HSM); the slip ratio threshold is not lower than the maximum value of the subcritical range.

[0019] Optionally, the execution of emergency intervention instructions includes at least one of the following: Send a pulse width modulation (PWM) block signal to the gate power amplifier of the inverter to reduce the three-phase drive current to a safe shutdown state and force the motor output torque to zero. The ESP active pressure build-up unit applies high-frequency pulse braking pressure to the target wheel that is experiencing longitudinal slippage; If the longitudinal slip ratio is not lower than the safety threshold, control of the MPC will not be restored; If the longitudinal slip ratio is lower than the safety threshold for N consecutive control cycles, the control of the HSM is released and the control of the MPC is restored. Based on the ramp function, the motor torque is increased from 0 to the output torque calculated by the MPC within a preset time.

[0020] Optionally, the method further includes: The execution results of monitoring and control commands include at least one of the following: actual wheel speed, actual motor response torque, and hydraulic cylinder pressure. The execution result is used as the current state of the vehicle and fed back to the MPC.

[0021] Optionally, the method further includes acquiring the multi-source sensing information through at least one of the following methods: Use an external temperature sensor to obtain the ambient temperature; The road surface type is determined using cameras and radar; The camera captures road surface texture features, water film, or snow and ice coverage. The real-time road adhesion coefficient is estimated based on the road texture features, the road type, and the water film or snow cover. Based on the current location of the vehicle, obtain meteorological information within a radius of a first distance centered on the current location; Read data from the battery management system; The wheel speed of at least one wheel is obtained using a wheel speed sensor; Obtain the driver's operation request.

[0022] Optionally, the method further includes: Align the multi-source sensing information in terms of both time and space dimensions; The time dimension alignment includes: determining a reference timestamp; obtaining the original data of the preceding and following frames adjacent to the target timestamp; and calculating the pre-time alignment data of the target timestamp based on the original data of the preceding and following frames using an interpolation method. The spatial dimension alignment includes: establishing a vehicle coordinate system; and projecting radar data and camera data into the vehicle coordinate system.

[0023] Optionally, identifying the current operating condition of the vehicle based on its multi-source perception information includes: The multi-source sensing information is input into a pre-trained working condition recognition model, and the temperature, adhesion coefficient and vehicle state are fuzzified and mapped to fuzzy sets; the fuzzy sets include the fuzzy types corresponding to the multi-source sensing information respectively; The fuzzy set is matched with the antecedent conditions of candidate rules in the rule base, the membership degree between the fuzzy type in the fuzzy set and the antecedent conditions is calculated, and the activation strength of the candidate rule is calculated based on the membership degree; the activation strength is used to indicate the degree of matching between the fuzzy set and the antecedent conditions. The target rule is determined based on the activation intensity of the candidate rule, and the working condition corresponding to the target rule is output as the current working condition.

[0024] Optionally, the method further includes: Activate the corresponding management sub-strategy based on the current operating condition; In the case of the current working condition being a complex condition of low temperature and low adhesion, the management sub-strategy includes: limiting the rise rate of the motor peak torque request to within a preset safety envelope; reducing the kinetic energy recovery intensity; sending a command to the chassis domain controller to increase the damping and controlling the ESP to actively establish the brake master cylinder pre-pressure.

[0025] Optionally, the method further includes: Based on the vehicle's current real-time current, battery state of charge (SOC), and internal resistance, the trajectory of terminal voltage changes in the future period is deduced. If the predicted terminal voltage in the terminal voltage change trajectory reaches the undervoltage lower limit, a power cap warning is sent to the vehicle control unit (VCU); the VCU is used to perform forward filtering on the accelerator pedal opening in response to the power cap warning.

[0026] Optionally, the method further includes: The system invokes a neural network model or a magic formula tire model to predict the vehicle's expected slip ratio based on the current expected torque. When the expected slip ratio reaches or exceeds the optimal slip ratio, a first-order inertial low-pass filter is superimposed on the torque request front end; wherein the time constant of the first-order inertial low-pass filter is dynamically adjusted according to the expected slip ratio.

[0027] According to another aspect of this application, an energy management device is provided, the device comprising: The identification module is used to identify the current operating condition of the vehicle based on the vehicle's multi-source perception information; The adjustment module is used to adjust the objective function of energy management and the constraints of vehicle control according to the current operating conditions. The objective function is used to balance the economy of energy management, battery temperature and driving safety. The generation module is used to generate control instructions based on the objective function and the constraints, and the control instructions are used to implement energy management that conforms to the objective function and the constraints.

[0028] Optionally, the objective function includes a weighted sum of energy consumption cost indicators, battery temperature maintenance indicators, and driving safety indicators; The adjustment module is used to adjust the energy consumption cost index corresponding to the energy consumption economic weight according to the current operating conditions. The adjustment module is used to adjust the battery temperature maintenance weight corresponding to the battery temperature maintenance index according to the current operating conditions. The adjustment module is used to adjust the driving safety weight of the driving safety index according to the current operating conditions.

[0029] Optionally, the initial value of the economic efficiency weight is a first value; the initial value of the battery temperature maintenance weight is a second value; and the initial value of the driving safety weight is a third value, wherein the first value is greater than the second value and the first value is greater than the third value. The adjustment module is configured to perform at least one of the following: When the current working condition is a complex working condition with low temperature and low adhesion, the energy consumption cost index corresponding to the energy consumption economy weight is reduced from the first value to the fourth value. When the current working condition is a complex working condition of low temperature and low adhesion, the weight of the battery temperature maintenance index is increased from the second value to the fifth value. When the current working condition is a complex working condition with low temperature and low adhesion, the driving safety weight corresponding to the driving safety index is increased from the third value to the sixth value. The fifth value is greater than the sixth value, and the sixth value is greater than the fourth value.

[0030] Optionally, the adjustment module is used to adjust the thermal management strategy according to the current operating conditions; and / or, Adjust torque distribution and anti-slip control according to the current operating conditions.

[0031] Optionally, the adjustment module is configured to execute at least one of the following thermal management strategies when the current operating condition is a complex condition of low temperature and low adhesion: Shut down the battery cooling circuit; Open the coolant series valve between the motor controller MCU and the battery pack; When the motor is not working, the positive temperature coefficient PTC heater is instructed to heat the coolant at maximum power. The target curve for battery temperature rise is adjusted to raise the battery temperature to a first temperature within a first duration; the first temperature is determined based on the limiting temperature of the battery charging power.

[0032] Optionally, the adjustment module is configured to perform a first-order low-pass filter on the requested torque corresponding to the accelerator pedal request when the current operating condition is a complex condition of low temperature and low adhesion; and to increase the system response time coefficient of the accelerator pedal: the time coefficient is used to control the speed at which the system responds to the accelerator pedal request; and / or, Under the current operating condition of low temperature and low adhesion, the maximum adhesion of the wheels is dynamically calculated based on the real-time estimated road adhesion coefficient; the output torque of the motor is limited based on the maximum adhesion and a safety factor; and / or, When the current working condition is a complex condition of low temperature and low adhesion, and the slip ratio of the first wheel is detected to be higher than the slip ratio threshold, the output torque of the first motor is reduced and the output torque of the second electrode is increased.

[0033] Optionally, the generation module is used to input the objective function and the constraints into the model predictive controller (MPC), and generate the control instructions based on the current state of the vehicle and the operation request; the predictive controller is used to generate a control sequence in the rolling time domain, the control sequence including at least one control instruction.

[0034] Optionally, the energy consumption cost index of the objective function includes a vehicle speed tracking deviation term; the battery temperature maintenance index of the objective function includes a thermal coupling cost term; the driving safety index of the objective function includes a desired acceleration deviation term and a motor torque increment penalty term; the objective function is used to evaluate the quality of control commands; the vehicle speed tracking deviation term is used to penalize the deviation between the actual vehicle speed and the target vehicle speed; the desired acceleration deviation term is used to penalize the deviation between the actual acceleration and the desired acceleration; the motor torque increment penalty term is used to penalize the change in motor torque within adjacent control cycles; the thermal coupling cost term is used to penalize the deviation between the actual power and the target power of the PTC heater. The generation module is used to iteratively solve the problem in the rolling time domain based on the current state of the vehicle and the operation request, according to the constraints and with the objective function as the goal, through the online quadratic programming solver of the MPC, and output the control sequence.

[0035] Optionally, the control commands include at least one of the following: target motor torque, PTC heating power, and ESP intervention threshold.

[0036] Optionally, the device further includes: The correction module is used to collect the vehicle's center of gravity speed at a first sampling frequency through the chassis domain controller and calculate the longitudinal slip ratio. The correction module is used to perform soft correction on the objective function and / or the constraint conditions in the MPC when the longitudinal slip ratio is in the subcritical range.

[0037] Optionally, the correction module is configured to perform at least one of the following: Reduce the torque increment step size; Adjust the weight of at least one of the following in the objective function: vehicle speed tracking deviation, thermal coupling cost, expected acceleration deviation, and motor torque increment penalty. Adjust the boundaries of the constraints.

[0038] Optionally, the correction module is used to increase the weight of the desired acceleration deviation term and decrease the weight of the vehicle speed tracking deviation term.

[0039] Optionally, the correction module is used to pre-reduce the maximum friction force of the virtual ground when the longitudinal slip ratio shows an upward trend; the maximum friction force of the virtual ground is used to constrain the maximum value of the longitudinal force of the wheel.

[0040] Optionally, the device further includes: The correction module is used to collect the vehicle center speed at a first sampling frequency through the chassis domain controller and calculate the longitudinal slip ratio. The correction module is configured to, when the longitudinal slip ratio is higher than the slip ratio threshold, not execute the control command output by the MPC, but execute an emergency intervention command through the hardware safety monitoring unit (HSM); the slip ratio threshold is not lower than the maximum value of the subcritical range.

[0041] Optionally, the correction module is configured to perform at least one of the following: Send a pulse width modulation (PWM) block signal to the gate power amplifier of the inverter to reduce the three-phase drive current to a safe shutdown state and force the motor output torque to zero. The ESP active pressure build-up unit applies high-frequency pulse braking pressure to the target wheel that is experiencing longitudinal slippage; If the longitudinal slip ratio is not lower than the safety threshold, control of the MPC will not be restored; If the longitudinal slip ratio is lower than the safety threshold for N consecutive control cycles, the control of the HSM is released and the control of the MPC is restored. Based on the ramp function, the motor torque is increased from 0 to the output torque calculated by the MPC within a preset time.

[0042] Optionally, the device further includes: The feedback module is used to monitor the execution results of control commands; the execution results include at least one of the following: actual wheel speed, actual motor response torque, and hydraulic cylinder pressure. The feedback module is used to feed back the execution result as the current state of the vehicle to the MPC.

[0043] Optionally, the acquisition module is configured to perform at least one of the following: Use an external temperature sensor to obtain the ambient temperature; The road surface type is determined using cameras and radar; The camera captures road surface texture features, water film, or snow and ice coverage. The real-time road adhesion coefficient is estimated based on the road texture features, the road type, and the water film or snow cover. Based on the current location of the vehicle, obtain meteorological information within a radius of a first distance centered on the current location; Read data from the battery management system; The wheel speed of at least one wheel is obtained using a wheel speed sensor; Obtain the driver's operation request.

[0044] Optionally, the acquisition module is used to align the multi-source sensing information in terms of time and space dimensions; The time dimension alignment includes: determining a reference timestamp; obtaining the original data of the preceding and following frames adjacent to the target timestamp; and calculating the pre-time alignment data of the target timestamp based on the original data of the preceding and following frames using an interpolation method. The spatial dimension alignment includes: establishing a vehicle coordinate system; and projecting radar data and camera data into the vehicle coordinate system.

[0045] Optionally, the recognition module is used to input the multi-source sensing information into a pre-trained working condition recognition model, and to fuzzify the temperature, adhesion coefficient and vehicle state, mapping them into fuzzy sets; the fuzzy sets include the fuzzy types corresponding to the multi-source sensing information respectively; The recognition module is used to match the fuzzy set with the antecedent conditions of candidate rules in the rule base, calculate the membership degree between the fuzzy type in the fuzzy set and the antecedent conditions, and calculate the activation intensity of the candidate rule based on the membership degree; the activation intensity is used to indicate the degree of matching between the fuzzy set and the antecedent conditions. The identification module is used to determine the target rule based on the activation intensity of the candidate rule, and output the working condition corresponding to the target rule as the current working condition.

[0046] Optionally, the identification module is used to activate the corresponding management sub-strategy based on the current operating condition; In the case of the current working condition being a complex condition of low temperature and low adhesion, the management sub-strategy includes: limiting the rise rate of the motor peak torque request to within a preset safety envelope; reducing the kinetic energy recovery intensity; sending a command to the chassis domain controller to increase the damping and controlling the ESP to actively establish the brake master cylinder pre-pressure.

[0047] Optionally, the identification module is used to predict the trajectory of terminal voltage changes in the future period based on the vehicle's current real-time current, battery charge state (SOC), and internal resistance. The identification module is used to send a power cap warning to the vehicle control unit (VCU) when the predicted terminal voltage in the terminal voltage change trajectory reaches the undervoltage lower limit; the VCU is used to perform forward filtering on the accelerator pedal opening in response to the power cap warning. Optionally, the identification module is used to call a neural network model or a magic formula tire model to predict the expected slip ratio of the vehicle based on the current expected torque; The identification module is used to superimpose a first-order inertial low-pass filter on the torque request front end when the expected slip ratio reaches or exceeds the optimal slip ratio; wherein the time constant of the first-order inertial low-pass filter is dynamically adjusted according to the expected slip ratio.

[0048] According to another aspect of this application, a computer device is provided, the computer device including a processor and a memory, the memory storing at least one instruction, at least one program, code set or instruction set, the at least one instruction, the at least one program, the code set or instruction set being loaded and executed by the processor to implement the energy management method as described above.

[0049] According to another aspect of this application, a computer-readable storage medium is provided, wherein at least one instruction, at least one program, code set, or instruction set is stored therein, wherein the at least one instruction, the at least one program, the code set, or the instruction set is loaded and executed by a processor to implement the energy management method described above.

[0050] According to another aspect of this application, a computer program product or computer program is provided, comprising computer instructions stored in a computer-readable storage medium. A processor of a computer device reads the computer instructions from the computer-readable storage medium and executes the computer instructions, causing the computer device to perform the energy management methods provided in various alternative implementations of the above aspects.

[0051] The beneficial effects of the technical solution provided in this application include at least the following: By acquiring multi-source sensing information, the system identifies the vehicle's current operating condition in real time, accurately determining the actual driving environment. Based on the vehicle's current operating condition, the objective function and constraints of energy management are adjusted, allowing the parameters of the energy management strategy to dynamically change according to the vehicle's actual driving state. This enables automatic adaptation to the optimal control strategy even under non-standard conditions such as extreme cold, low traction, and congestion, improving the rationality of energy regulation and its generalization ability. Furthermore, the objective function balances the three major objectives of energy management: economy, battery temperature, and driving safety. It adjusts the weights of each objective according to the current operating condition, coordinating their priorities and avoiding the continued use of energy-saving strategies under normal operating conditions when battery temperature is too high or road traction is extremely low, effectively resolving the conflict and coupling problems between multiple objectives. Attached Figure Description

[0052] To more clearly illustrate the technical solutions in the embodiments of this application, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0053] Figure 1 This is a schematic diagram of a controller provided in an exemplary embodiment of this application; Figure 2 This is a schematic diagram of an energy management method provided in an exemplary embodiment of this application; Figure 3 This is a schematic diagram of an energy management method provided in an exemplary embodiment of this application; Figure 4 This is a schematic diagram of an energy management method provided in an exemplary embodiment of this application; Figure 5 This is a flowchart of an exemplary embodiment of the energy management method provided in this application; Figure 6 This is a flowchart of an exemplary embodiment of the energy management method provided in this application; Figure 7 This is a schematic diagram of the structure of an energy management device provided in an exemplary embodiment of this application; Figure 8 This is a schematic diagram of the structure of an in-vehicle terminal provided in an exemplary embodiment of this application.

[0054] The accompanying drawings, which are incorporated in and form part of this specification, illustrate embodiments consistent with this application and, together with the description, serve to explain the principles of this application. Detailed Implementation

[0055] To make the technical solution and advantages of this application clearer, the embodiments of this application will be described in further detail below.

[0056] The terms "first," "second," "third," and "fourth," etc., used in the specification, claims, and accompanying drawings of this application are used to distinguish different objects, not to describe a specific order. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover non-exclusive inclusion. For example, a process, method, system, product, or apparatus that includes a series of steps or units is not limited to the listed steps or units, but may optionally include steps or units not listed, or may optionally include other steps or units inherent to these processes, methods, products, or apparatuses.

[0057] It should be noted that the information (including but not limited to user device information, user personal information, etc.), data (including but not limited to data used for analysis, data stored, data displayed, etc.) and signals involved in this application are all authorized by the user or fully authorized by all parties, and the collection, use and processing of related data must comply with the relevant laws, regulations and standards of the relevant countries and regions.

[0058] Figure 1 A block diagram of a controller 110 provided in an exemplary embodiment of this application is shown.

[0059] The controller 110 is installed in the vehicle and is used to control the vehicle. The controller 110 has an energy management program installed and running. The energy management program can run to implement the energy management method provided in the embodiments of this application.

[0060] The controller 110 includes a first memory and a first processor. The first memory stores an energy management program; the energy management program is invoked and executed by the first processor to implement the energy management method provided in this application. The first memory may include, but is not limited to, the following: Random Access Memory (RAM), Read Only Memory (ROM), Programmable Read-Only Memory (PROM), Erasable Programmable Read-Only Memory (EPROM), and Electrically Erasable Programmable Read-Only Memory (EEPROM).

[0061] The first processor can consist of one or more integrated circuit chips. Optionally, the first processor can be a general-purpose processor, such as a central processing unit (CPU) or a network processor (NP). Optionally, the first processor can implement the energy management method provided in this application by running programs or code.

[0062] For example, controller 110 includes an EMS (Energy Management System) controller. The EMS controller is used to acquire multi-source perception information of the vehicle, identify the current operating conditions, and reconstruct the objective function and constraints.

[0063] For example, controller 110 includes MPC (Model Predictive Control). MPC is used to solve for control commands based on the objective function and constraints. MPC is also used to adjust the generator target torque, PTC (Positive Temperature Coefficient) heating power, and ESP (Electronic Stability Program) engagement threshold based on the control commands.

[0064] For example, the controllers involved in the embodiments of this application also include chassis domain controllers, VCU (Vehicle Control Unit), HSM (Hardware Security Module), ESP, and BMS (Battery Management System).

[0065] The chassis domain controller calculates the vehicle's center of gravity speed and real-time longitudinal slip ratio using wheel speed sensors and an INS (Inertial Navigation System), providing chassis status feedback to the upper-level control strategy. The chassis domain controller is responsible for vehicle dynamic control and driving attitude, directly determining handling safety, driving stability, and ride comfort. It integrates and coordinates the drivetrain, steering, and braking systems.

[0066] The VCU receives power cap warnings from the EMS / MPC and performs forward filtering on the accelerator pedal opening. The VCU is responsible for collecting and analyzing the driver's intentions in real time, such as acceleration, braking, and gear shifting requests, and combining this with the status information of various vehicle components to calculate the optimal torque and power commands.

[0067] HSM is used to bypass MPC output when the slip ratio exceeds the critical threshold of 15%, triggering the highest priority emergency intervention, forcing the motor output torque to zero, and cooperating with ESP to perform hydraulic braking in conjunction with intermittent braking.

[0068] ESP receives EMS commands and actively builds up pre-stress in the master cylinder, shortening response time. It also applies high-frequency pulse braking pressure to slipping wheels when the highest priority emergency intervention is triggered. By continuously monitoring the vehicle's driving status (such as wheel speed, steering wheel angle, lateral acceleration, etc.), ESP actively brakes one or more wheels and adjusts engine power output when it detects a tendency towards understeer or oversteer, helping the vehicle drive as intended by the driver, preventing skidding and fishtailing, and thus maintaining vehicle stability.

[0069] The Battery Management System (BMS) provides data such as average battery pack temperature, State of Charge (SOC), and internal resistance to help the system determine vehicle status. The BMS also sets the undervoltage protection lower limit; when the MPC (Maximum Power Controller) prediction voltage reaches this limit, a power cap warning is triggered to protect battery safety. The BMS can monitor the voltage, current, and temperature of each cell in the battery pack in real time to prevent dangerous situations such as overcharging, over-discharging, and thermal runaway. It also communicates with the Vehicle Control Unit (VCU) via the CAN (Controller Area Network) bus to report battery status.

[0070] This application provides an energy management method based on complex environmental conditions. It can dynamically match the optimal energy allocation and safety protection strategies based on real-time environmental and vehicle status inputs, thereby achieving energy consumption optimization and system robustness enhancement under all operating conditions while ensuring driving safety.

[0071] Figure 2 This is a flowchart illustrating an exemplary embodiment of an energy management method provided in this application. The method can be implemented by, for example... Figure 1 The controller shown executes this method, which includes the following steps.

[0072] For example, this embodiment only details the complex extreme working condition of extremely low ambient temperature (-15℃) and extremely poor road surface adhesion coefficient (μ<0.3). However, its control framework "multi-source perception → fuzzy rule matching → target reconstruction → predictive power intervention" has the ability to generalize working conditions. By adjusting the membership threshold and calibration parameters of fuzzy sets such as temperature and adhesion coefficient, the control framework can also be transferred to other extreme environments such as high temperature and high altitude, so as to achieve cross-working condition coordination of energy consumption and safety.

[0073] Step 201: High-frequency acquisition and fusion sensing of multi-source sensing information.

[0074] EMS collects multi-dimensional perception data in real time at a frequency of 100Hz through the vehicle's CAN bus and onboard sensor network, and performs time and spatial alignment of perception data from different sensors. In the time dimension, interpolation is used to unify the perception data from different sensors to the same timestamp. In the spatial dimension, the radar and camera-perceived targets are mapped to the same vehicle coordinate system through coordinate transformation, ensuring that the multi-source perception data has comparability and a basis for fusion in time and space.

[0075] For example, such as Figure 3 As shown, step 201 may include steps 201-1 to 201-5 as follows.

[0076] Step 201-1: High-frequency acquisition of multi-source sensing information.

[0077] Call an external temperature sensor (such as a temperature sensor located under the front bumper or rearview mirror) to confirm the ambient temperature. For example, if the ambient temperature T < -15℃ (at this temperature, the water film or snow on the road surface is difficult to melt and will form an extremely hard and slippery ice layer, causing the adhesion coefficient to drop sharply).

[0078] The roadbed surface type is determined by combining optical cameras with millimeter-wave radar image recognition algorithms. Since dry asphalt, wet asphalt, and concrete reflect laser light differently, the road surface type can be inferred by analyzing the reflection intensity distribution.

[0079] Analyze road surface texture features and water / snow film coverage. Using the Burckhardt tire-road adhesion model, and incorporating road surface texture features, road surface medium, water film, and other conditions, estimate the real-time road surface adhesion coefficient, for example, adhesion coefficient μ < 0.3.

[0080] The vehicle obtains a weather warning within a three-kilometer radius of the vehicle via a V2X (Vehicle to Everything) communication module, confirming that there is no warming trend that could cause the ice to melt within the next 30 minutes.

[0081] Step 201-2: Time Alignment. Standardize timestamps and perform interpolation compensation.

[0082] Cameras typically output image frames at 30Hz or 60Hz, millimeter-wave radars often operate at 20Hz or 25Hz, while lidar may only reach 10Hz. This means that within one second, the data generated by different sensors can occur at completely different points in time. Even if the frequencies are integer multiples (such as 30Hz and 10Hz), their exposure / scan start times are almost impossible to perfectly coincide. For example, a camera might expose at 0ms, while a radar might only output a point cloud at 16ms. Without time alignment, at high vehicle speeds (e.g., 30m / s, or 108km / h), a mere 50ms time difference can cause the target position to drift by 1.5 meters. Directly fusing unaligned "past targets" with "current targets" can lead to "ghosting" or "phantom targets" in the perception system, severely interfering with decision-making (e.g., false triggering of AEB (Automatic Emergency Braking)). Therefore, time alignment of multi-dimensional perception data is achieved through the following method: (1) Establish a unified reference timestamp (master axis). Since visual images are "snapshots" and require the highest time accuracy, while radar and lidar point clouds are "scans" and have a slightly larger time span, the camera exposure start time (Mid-Exposure) is used as the global master timestamp.

[0083] (2) Select the nearest original data (previous and next frames). Assume the unified reference time is T. target(For example, at 50ms of camera exposure). If the latest radar frame is in T1 (e.g., 40ms) and the next radar frame is in T2 (e.g., 80ms), the system will cache the radar-sensed target (including position x, y, z and velocity V) at both times. x V y ).

[0084] (3) Perform "interpolation" calculation. For radar targets, the system does not directly take the measured values ​​of T1 and T2, but assumes that the target is moving in uniform linear motion within a very short time interval. Using the position and velocity changes at times T1 and T2, linear interpolation is performed to obtain T. target The most likely precise location at any given moment. For the camera (if it is necessary to map image pixels to vehicle coordinates), the vehicle's own motion (IMU (Inertial Measurement Unit) + wheel speedometer) before and after this frame is also used for compensation to deduce the "true" projected position of the object in the image at the current reference moment.

[0085] Step 201-3: Spatial Dimension Alignment. Perform coordinate transformation.

[0086] Because the radar and camera are installed in different locations on the vehicle (the radar may be mounted on the bumper, and the camera may be mounted near the rearview mirror), each establishes a measurement coordinate system with its own physical center as the origin. The radar directly outputs the target's distance, angle, and Doppler velocity in the polar coordinate system; the camera outputs a two-dimensional image (u,v) in the pixel coordinate system and a bounding box inferred by deep learning. To determine whether the point detected by the radar and the vehicle captured by the camera are the same object, they must be projected from their respective local coordinate systems to a unified vehicle coordinate system.

[0087] (1) Sensor local coordinates to three-dimensional spatial coordinates. The polar coordinates (range r, azimuth α) measured by the radar are converted into three-dimensional Cartesian coordinates (x, y, y) in the radar local coordinate system by trigonometric functions. r ,y r ,z r If the camera has ranging capabilities (such as binoculars or LiDAR assistance), then using the pinhole camera inverse projection model, the image pixel coordinates (u,v) are combined with depth estimation to be inversely projected into a three-dimensional point (x,v) in the camera coordinate system. c ,y c ,z c ).

[0088] (2) Extrinsic parameter matrix (rotation matrix R + translation vector t) mapping. Through prior joint calibration, the rotation angles (roll, pitch, yaw) and three-dimensional spatial offsets of the radar and camera relative to the rear axle center (or center of mass) of the vehicle are obtained. Subsequently, the three-dimensional coordinate points of the two sensors are multiplied by the corresponding extrinsic parameter matrices and projected uniformly onto the vehicle coordinate system (defined: X-axis forward, Y-axis left, Z-axis upward).

[0089] Step 201-4: Vehicle status analysis.

[0090] Read data from the Battery Management System (BMS), such as battery temperature, State of Charge (SOC), and internal resistance. For example, the average battery pack temperature T is maintained at -12°C, and the SOC is approximately 50%. When the vehicle is in an extremely cold environment without active heating, the internal resistance R increases by about 40% compared to normal temperature.

[0091] Determine the driver's intention. For example, if the accelerator pedal opening α is less than 15% and the rate of pedal change is relatively small, it is judged as a smooth start-up phase.

[0092] Read chassis data. For example, read the wheel speeds displayed by the wheel speed sensors. If the difference in wheel speeds exceeds 5%-15%, there may be a risk of slippage.

[0093] Step 201-5: Output the multi-source perception information fusion result. The multi-dimensional perception data, after being aligned in time and space, and the vehicle status data are fused to obtain multi-source perception information.

[0094] Step 202: Accurate identification of working conditions based on fuzzy reasoning.

[0095] (1) such as Figure 4 As shown, the multidimensional perception information obtained by the above fusion is input into the pre-trained working condition recognition model.

[0096] .

[0097] in, , , These represent temperature, coefficient of adhesion, and vehicle status (such as starting, acceleration, and braking), respectively. The logical AND operator is used; Defuzzify is for defuzzification; and n is the total number of rules. The output of the i-th rule is to sense changes in the state of the sensing device.

[0098] Input variables are fuzzified. Temperature (low, medium, high), adhesion coefficient (extremely low, low, medium, high), and battery temperature rise rate (slow, fast) are mapped to fuzzy sets.

[0099] (2) Perform rule base matching. For example, IF(T< 15℃) AND (μ<0.3) AND (Vehicle_State==Start) THEN Mode=Extreme_Cold_Low Grip, that is, if the battery temperature is less than 15℃ and the adhesion coefficient is less than 0.3 and the vehicle state is in the starting state, then as follows Figure 4 As shown, the current vehicle condition is locked as a complex working condition with low temperature and low adhesion.

[0100] After the fused features are input, the inference engine calculates the firing strength of all rules in parallel, that is, it calculates the truth value of the antecedent conditions using the min or product operators. For example, in the complex condition of low temperature and low adhesion, the antecedent conditions are that the battery temperature is less than 15°C, the adhesion coefficient is less than 0.3, and the vehicle is in a starting state. If the current temperature T = At 18℃, with an adhesion coefficient μ=0.25, and the vehicle in a starting state (Start) and the sensor self-test passed, the activation strength of this rule will reach approximately 1.0 (perfect match), suppressing other rules with low matching degrees in rule competition. Subsequently, the system uses the centroid defuzzification method to solve each output fuzzy set into a clear discrete working condition pattern code, and locks the final discrimination result through a confidence threshold.

[0101] (3) such as Figure 4 As shown, a safety-priority control strategy is issued. Once the inference engine determines that the confidence level meets the standard, the system officially locks the current condition as "low temperature, low adhesion complex operating condition" and activates the corresponding safety-priority energy management sub-strategy. Its specific execution constraints include: 1) Drive torque gradient limit: Limit the rise slope of the motor peak torque request within a preset safety envelope (e.g., <50Nm / 100ms) to prevent the instantaneous burst force of the motor from exceeding the static friction limit under extremely low adhesion coefficient when the driver presses the accelerator hard.

[0102] 2) Regenerative braking weight reduction: Reduce the kinetic energy recovery intensity (from -0.3g in standard mode to -0.1g), prioritize hydraulic mechanical braking, and avoid excessive single-axle recovery torque that could cause rear wheel lock-up or yaw instability.

[0103] 3) Active suspension and ESP pre-boost: Sends commands to the chassis domain controller to increase the damper damping to suppress pitch and commands the Electronic Stability Program (ESP) to actively build up brake master cylinder pre-pressure, shortening the response time by about 200ms.

[0104] (4) such as Figure 4 As shown, forward prediction is performed based on the physical model, and the output is smoothly controlled.

[0105] 1) Battery usable power degradation prediction. Based on the current real-time current, battery SOC (state of charge), and estimated internal resistance (R). int (The voltage increases exponentially with decreasing temperature). The Extended Kalman Filter (EKF) is used to predict the terminal voltage change trajectory over the next 10 seconds (corresponding to approximately 100 control cycles). When the predicted terminal voltage is about to reach the undervoltage protection lower limit set by the BMS, the system issues a "power cap warning" to the VCU in advance (e.g., informing the system that the maximum output power will be limited to 70% of the current value in the 8th second). This allows the VCU to perform forward filtering on the accelerator pedal opening, avoiding sudden "cliff-like" power reduction.

[0106] 2) Wheel transient slip ratio suppression. For roads with extremely low traction, tire forces exhibit strong nonlinear characteristics. Using a pre-trained neural network or a simplified Pacejka Magic Formula tire model, the algorithm predicts that under the driver's current expected torque, the difference between wheel speed and vehicle speed (i.e., slip ratio) will rapidly increase to above the optimal slip ratio (approximately 15%~20%) within the next 1-2 seconds. To avoid severe jerking (excessive longitudinal impact jerk) caused by sudden power drops, the algorithm pre-loads a first-order inertial low-pass filter (with time constant τ dynamically adjusted according to slip ratio risk) at the torque request front end. This transforms the steep torque drop into a gently sloping ramp, ensuring the wheels do not lock up while controlling the longitudinal impact of the vehicle body to <2m / s². 3 Within the comfort threshold, it creates a smooth and safe driving experience for the driver on extremely cold and icy roads.

[0107] Step 203: Multi-objective dynamic reconstruction and adaptive adjustment of strategy parameters.

[0108] For the locked operating condition, the EMS dynamic reconfiguration optimization objective function J and its constraints are adjusted as follows: 1. Activation of thermal management strategy.

[0109] The waste heat recovery logic immediately shuts down the battery cooling circuit and opens the liquid-cooled series valve between the motor controller (Microcontroller Unit, MCU) and the battery pack.

[0110] If the motor is not working, the PTC (positive temperature coefficient) heater is instructed to engage at maximum power (limited by the low-voltage battery capacity).

[0111] Target setting: Set a target curve for rapid battery temperature rise, requiring the cell temperature to be raised above 0°C within 300 seconds to remove charging power limitations and reduce internal resistance.

[0112] Power allocation involves allocating a fixed percentage (e.g., 2kW) from the drive power budget specifically for the thermal management system, which means sacrificing some power response to supply battery thermal management.

[0113] 2. Torque distribution and anti-slip control.

[0114] The torque required by the driver T req Based on this, a first-order low-pass filter is introduced.

[0115] The time constant τ for the operation request response is increased from the usual 0.1s to 0.5s to eliminate torque mutations and prevent instantaneous impacts from exceeding the road surface adhesion limit.

[0116] Based on the real-time estimated road adhesion coefficient μ, the maximum adhesion force F of each wheel is dynamically calculated. max =μ·F z (F) z (for vertical loads), and limit the motor output torque to T. limit =F max ·r wheel / i gear ·K safe (T) limit F is the output torque limit value of the motor. max For the maximum adhesion of the wheel, r wheel Let i be the radius of the wheel. gear K is the gear ratio. safe For safety, we take a factor of 0.8.

[0117] If an abnormal increase in the speed of a single wheel is detected (slip ratio S>5%), the motor on that side will immediately be subject to millisecond-level torque reduction, and the torque vector will be transferred to the side with better traction (if it is a dual-motor vehicle).

[0118] 3. Optimize the weight reconstruction of the objective function.

[0119] The original objective function J=w1E cost The weighting coefficients in +w2Comfort and +w3Safety are dynamically adjusted.

[0120] W1 (energy consumption economic weight) has been reduced from 0.6 to 0.1. Under complex operating conditions of low temperature and low adhesion, the pursuit of extreme energy saving is no longer pursued. Appropriate increases in energy consumption are allowed in exchange for safety and battery activity. A small amount of energy consumption is used to avoid the high rescue costs and inefficiency caused by vehicle shutdown.

[0121] w2 (battery temperature maintenance weight) is increased from 0.2 to 0.5. This forces the system to prioritize maintaining the battery's thermal condition.

[0122] The weight of w3 (driving safety / stability) has been increased from 0.2 to 0.4. Anti-skid and anti-loss-of-control measures have been made the core assessment indicators.

[0123] E cost The objective function J represents the energy consumption cost index, Comfort is the rate of change of longitudinal acceleration, and Safety is the sum of the squares of the ratios of longitudinal force to ultimate adhesion force of each wheel, representing the driving safety index. It should be noted that E in the objective function J... cost Comfort and Safety are all normalized or standardized values. E cost The original values ​​of Comfort and Safety are uniformly scaled to the range (0,1] to obtain normalized values, or E is... cost The standardized values ​​are obtained by subtracting the mean from the original values ​​of Comfort and Safety and dividing by the standard deviation.

[0124] Step 204: Model Predictive Control (MPC) Execution and Real-Time Feedback Correction.

[0125] 1. Rolling time-domain optimization and the issuance and execution of optimal control instructions.

[0126] After completing the working condition mode locking (such as "low temperature and low adhesion complex working condition"), the system inputs the improved multi-objective cost function and hard constraint boundary into the model predictive controller (MPC).

[0127] MPC employs explicit MPC (eMPC) combined with an online QP (Quadratic Programming) solver, operating within a finite rolling time domain ( The optimal control sequence is generated within 20 steps.

[0128] (1) Time domain parameter setting: control step size Δt = 100ms, therefore the total prediction time domain is The time domain is 0.5 Hz, which precisely covers the main response bandwidth of the vehicle's longitudinal dynamics. This ensures sufficient prediction while avoiding the cumulative error caused by model mismatch due to an excessively long time domain.

[0129] (2) Improved objective function It not only includes the traditional tracking error term (vehicle speed tracking deviation) Deviation from expected acceleration It also introduced a control increment penalty term () ) and thermal coupling cost term:

[0130] in, Assigning weights to changes in vehicle speed As the weight for acceleration change, Weighting for changes in motor torque. The weight of the PTC heating function. The weights are significantly increased to forcefully suppress drastic torque fluctuations and ensure that longitudinal jerk on low-friction surfaces does not exceed the comfort threshold. This represents the number of steps in the rolling time domain. Let be the vehicle speed tracking deviation at step k. Let be the expected acceleration deviation at step k. This is the motor torque increment (control increment penalty term). For current thermal management power loss / consumption, The target thermal management reference power is [value]. It should be noted that the objective function [function name]... In , , , , All values ​​are normalized or standardized. , , , , The original values ​​are uniformly scaled to the interval (0,1] to obtain normalized values, or, the original values ​​are scaled to the interval (0,1] to obtain normalized values. , , , , The original value minus the mean and divided by the standard deviation yields the standardized value.

[0131] (3) Real-time optimal instruction output: After solver iteration, MPC only issues the first control quantity in the instruction sequence at the current moment, including: 1) Target torque of the motor : A smooth torque request after gradient limiting, rather than the driver's original pedal mapping value; 2) PTC heating power: Based on the current battery temperature and the predicted available power in the next 10 seconds, the duty cycle of the positive temperature coefficient (PTC) heater is adjusted in advance to actively heat up the battery pack to restore electrochemical activity and prevent subsequent power drop. 3) ESP intervention threshold: dynamically raise or lower the yaw rate error threshold of the electronic stability program (for example, tighten the intervention threshold from the usual ±5° / s to ±3° / s in low-adjustment mode) to allow the vehicle stability control to intervene in advance and prevent problems before they occur.

[0132] 2. Real-time slip ratio calculation and soft correction (slip ratio in the range of 5%~15%).

[0133] The MPC command does not terminate the open loop; instead, it enters the high-speed feedback correction loop. The chassis domain controller acquires data from the wheel speed sensors in real time at a sampling frequency of 1 kHz (every 1 ms). ) and the vehicle's center of gravity speed calculated by the INS inertial navigation system ( ), and calculate the real-time longitudinal slip ratio:

[0134] in A protection coefficient is used to prevent the denominator from being zero during static start-up.

[0135] When the slip ratio S is in the "subcritical range" of [5%, 15%], the system determines that the tire has entered the edge of the nonlinear adhesion region, but is still within a controllable and stable slip range. At this point, MPC does not intervene aggressively, but instead initiates online weight self-tuning soft correction: (1) Reduce the torque increment step size ( The boundary constraints of the QP solver are tightened, reducing the maximum allowable torque rise / fall in each control cycle from the usual 200 Nm to 50 Nm. This is equivalent to applying a first-order low-pass filter to the accelerator pedal request, thus slowing down the power release rate.

[0136] (2) Dynamic transfer of objective function weights: temporarily increasing the weights in the objective function (Battery temperature maintenance weight) and reduce (Energy efficiency weighting) The system actively sacrifices some vehicle speed tracking accuracy in exchange for absolute stability of longitudinal acceleration.

[0137] (3) Rolling time domain contraction: If the slip ratio is on the rise (i.e. S>0), the prediction model inside the MPC will adjust the virtual ground maximum friction constraint downward in advance, so that the torque command in the control sequence of the next few steps will automatically have attenuation feedforward, realizing the smooth domestication of "pre-decrease before exceeding the limit".

[0138] 3. Hard security monitoring layer triggers and emergency hard-coded rollback (S>15%).

[0139] When the slip ratio exceeds the critical threshold of 15%, it means that the tire has completely exceeded the adhesion ellipse limit and entered a state of deep slippage / lock-up. At this point, the linear tire model is completely invalid, and the mathematical basis of MPC optimization based on the nominal model no longer holds (the model mismatch is severe, and the calculated torque may oscillate and diverge). The system's built-in independent hardware safety monitoring unit (HSM) will immediately bypass the MPC software output and trigger the highest priority emergency intervention mode. The emergency intervention mode includes: (1) Torque to zero under 10ms extreme response: The safety monitoring logic directly sends a PWM shutdown signal (Shut-down PWM) to the inverter's GateDriver without going through the CAN bus delay (usually only 2-4ms). Within ≤10ms, the three-phase drive current is reduced to a safe shutdown state, forcing the motor output torque to zero. This completely cuts off the power source and prevents the slip ratio from diverging to 100% idling due to continuous drive.

[0140] (2) Hydraulic braking combined with chirp braking: While cutting off power, the ESP active pressure building unit responds immediately, applying high-frequency pulsating braking pressure (commonly known as "chirp braking") at a frequency of 15~20Hz to the target wheel (or multiple wheels) that has slipped. This high-frequency pressure fluctuation can effectively break the static friction sticking state between the tire and the ice surface, and in conjunction with the dynamic recovery of the ground adhesion coefficient, quickly pull the wheel speed back to near the reference speed.

[0141] (3) Hysteresis Recovery Logic and MPC Smooth Restart: Emergency intervention will continue until the slip ratio is forcibly reduced and stabilized within a safe range (S<5%). To avoid control chatter (frequent switching of the system at threshold boundaries), the recovery logic uses a hysteresis comparator—the HSM releases control and removes torque lock-up only after S is below 5% for three consecutive control cycles. Subsequently, the system will not instantly switch back to the MPC command, but will use a ramp function to gradually increase the motor torque from 0 to the requested value calculated by the current MPC within 200ms, achieving a seamless transition between hard backoff and soft control.

[0142] 4. Closed-loop iteration and full-link redundancy security assurance.

[0143] Each feedback correction and actual execution result (including actual wheel speed, actual motor response torque, and hydraulic cylinder pressure) will be fed back to the input of the MPC prediction model as a new initial state, covering the prediction error of the previous moment, forming a tight closed loop of "prediction-execution-feedback-correction".

[0144] This layered architecture of "MPC routine fine control + hard-line safety monitoring with unconditional priority" not only ensures driving smoothness and economy for 95% of the time under harsh conditions, but also provides an insurmountable functional safety baseline (meeting ASIL-D (Automotive Safety Integrity Level D) requirements) when extreme physical limits are triggered, ensuring that every longitudinal impact on low-temperature, low-adhesion road surfaces is smooth and controllable, and that every extreme sideslip can be transiently eliminated.

[0145] In summary, the method provided in this application constructs a closed-loop control system of "perception-prediction-decision-correction". It not only collects in-vehicle signals such as vehicle speed and SOC, but also integrates high-precision maps, meteorological data, and traffic information to comprehensively depict the "complex environment" of vehicle operation. Artificial intelligence algorithms are used to identify current operating conditions, such as specific conditions like "cold start in low temperatures," "continuous uphill climbing," or "frequent start-stop," and predict future trends. Based on the characteristics of different operating conditions, the weights of fuel consumption, energy consumption, and SOH (State of Health, battery health) in the objective function are dynamically adjusted. For example, in extremely cold conditions, the priority of battery insulation is increased; in long downhill conditions, kinetic energy recovery efficiency is maximized. The model deviation is continuously corrected through actual driving data, giving the strategy the characteristic of "becoming more accurate with use."

[0146] The method provided in this application significantly improves the energy utilization efficiency of vehicles under non-standard operating conditions such as complex and variable road conditions (e.g., mountain climbing, urban congestion) and extreme climates (e.g., extreme cold, high temperature) through deep optimization of energy management strategies and precise control of intelligent algorithms. It is expected to reduce overall energy consumption by 5% to 10%. At the same time, its intelligent temperature control management system and pulse charging protection mechanism can monitor the cell status in real time and dynamically adjust charging and discharging parameters, effectively mitigating irreversible damage to the power battery caused by extreme environments such as high-temperature thermal runaway or low-temperature lithium plating, thereby significantly extending the service life of the battery pack. In addition, through the fine tuning of the motor torque output characteristics and the synergistic optimization of the vehicle control strategy, the vehicle's power response is smoother and more linear, completely eliminating the jerking sensation of traditional models during start-stop or acceleration and deceleration, and significantly enhancing the driving feel and overall ride comfort.

[0147] Figure 5 This is a flowchart illustrating an exemplary embodiment of an energy management method provided in this application. The method can be implemented by, for example... Figure 1 The controller shown executes this method, which includes the following steps.

[0148] Step 520: Identify the vehicle's current operating condition based on the vehicle's multi-source perception information.

[0149] For example, multi-source perception information includes at least one of environmental perception information, vehicle status information, and chassis domain controller feedback information.

[0150] The environmental perception information includes at least one of the following: ambient temperature, road surface images captured by optical cameras, environmental point clouds identified by radar, road surface adhesion coefficient, and weather warning information.

[0151] Vehicle status information includes at least one of the following: battery management system data (such as average battery pack temperature, SOC, internal resistance, etc.), accelerator pedal opening, accelerator pedal change rate, and wheel speed difference.

[0152] The chassis domain controller feedback information includes at least one of the following: wheel linear velocity obtained by wheel speed sensors, and vehicle center of gravity speed calculated by the INS inertial navigation system.

[0153] For example, after the vehicle is started, the EMS periodically acquires multi-source perception information of the vehicle according to a preset sampling period.

[0154] For example, EMS collects multi-dimensional perception information in real time at a preset frequency (e.g., 100Hz) through the vehicle's CAN bus and on-board sensor network, and performs spatiotemporal alignment and fusion of data from different sensors. In the time dimension, it unifies the data to the same reference timestamp through interpolation, and in the spatial dimension, it maps the target coordinates of radar and camera to the same vehicle coordinate system through the external parameter matrix obtained by joint calibration, ensuring that the multi-source information has comparability and a basis for fusion.

[0155] For example, based on the acquired multi-source perception information, a fuzzy inference algorithm is used to determine the specific operating condition of the vehicle. For instance, the current operating condition may include: low temperature condition, low adhesion condition, low temperature and low adhesion condition, normal condition, high temperature condition, high altitude condition, high adhesion condition, etc. The type of operating condition can be categorized as needed, and this embodiment does not impose any limitations on this.

[0156] Step 530: Adjust the objective function of energy management and the constraints of vehicle control according to the current operating conditions. The objective function is used to balance the economy of energy management, battery temperature and driving safety.

[0157] The objective function is a mathematical function used by EMS to quantify the priority and optimization direction of multiple control objectives. The objective function integrates multiple control objectives into a single cost function through a weighted summation, and MPC generates the optimal control sequence by minimizing this cost function.

[0158] For example, the objective function includes economic indicators, battery temperature indicators, and driving safety indicators, used to balance the economic objectives, battery temperature objectives, and driving safety objectives.

[0159] Because the requirements for economy, battery temperature and driving safety vary under different operating conditions, in order to make the vehicle's energy distribution strategy more suitable for the current operating conditions, the EMS will reconstruct the weights of each indicator in the objective function and adjust the energy distribution strategy.

[0160] Constraints are the physical limits, safety boundaries, and hardware limitations that must be satisfied when generating control instructions. Constraints define the solvable range of control instructions, ensuring that they do not exceed the actuator's capabilities or introduce safety risks.

[0161] For example, the constraints may include at least one of the following: motor torque constraint, slip ratio constraint, undervoltage protection lower limit, PTC heating power constraint, ESP intervention threshold constraint, and impact constraint.

[0162] Motor torque constraints can include: constraining the upward slope of output torque, constraining the downward slope of output torque, constraining the value of output torque, constraining the magnitude of torque decrease, constraining the magnitude of torque increase, and constraining output torque to zero.

[0163] Slip ratio constraints can include: setting a subcritical slip ratio threshold, setting a critical slip ratio threshold, and setting a normal slip ratio threshold.

[0164] The undervoltage protection lower limit is the minimum allowable voltage threshold set by the BMS to protect the battery from damage due to over-discharge. When the actual voltage falls below the undervoltage protection lower limit, the system will automatically cut off the power supply or stop working to terminate the battery discharge.

[0165] PTC heating power constraint is used to constrain the output power of PTC heaters.

[0166] ESP intervention threshold constraints are used to adjust the trigger thresholds for ESP intervention, such as yaw rate error threshold, steering wheel angle threshold (after the steering wheel angle exceeds the threshold, the matching degree between steering angle and vehicle speed is monitored), sideslip angle and sideslip rate threshold, etc.

[0167] Impact constraints are used to constrain the longitudinal impact of a vehicle.

[0168] Step 540: Generate control commands based on the objective function and constraints. The control commands are used to achieve energy management that meets the objective function and constraints.

[0169] After obtaining the objective function and constraints that match the current operating conditions, control commands are generated based on the objective function, constraints, and the current state of the vehicle, according to the driver's operation request, so that the actuators execute the control commands to achieve energy management that matches the current operating conditions.

[0170] In summary, the method provided in this application acquires multi-source sensing information and identifies the vehicle's current operating condition in real time based on this information, accurately determining the vehicle's actual driving environment. By adjusting the objective function and constraints of energy management based on the vehicle's current operating condition, the parameters of the energy management strategy can dynamically change with the vehicle's actual driving state. This allows for automatic adaptation to the optimal control strategy even under non-standard conditions such as extreme cold, low traction, and congestion, improving the rationality of energy regulation and its generalization ability to different operating conditions. Furthermore, the objective function can balance the three major objectives of energy management: economy, battery temperature, and driving safety. By reconstructing the weights of each objective based on the current operating condition and coordinating their priorities, it avoids using energy-saving strategies under normal operating conditions when the battery temperature is too high or the road surface traction is extremely low, effectively resolving the conflict and coupling problem between multiple objectives.

[0171] Figure 6 This is a schematic flowchart illustrating an exemplary embodiment of an energy management method provided in this application. This method can be used for, for example... Figure 1 The controller shown. Based on Figure 5 In the illustrated embodiment, step 510 may be included before step 520; and / or, step 530 may include step 531; and / or, step 540 may include step 541; and / or, step 542 to step 546 may be included after step 540.

[0172] Step 510: Obtain multi-source perception information of the vehicle.

[0173] For example, acquiring multi-source sensing information includes at least one of the following: 1) Use an external temperature sensor to obtain the ambient temperature.

[0174] For example, ambient temperature data can be collected in real time using an external temperature sensor installed under the vehicle's front bumper or rearview mirror. The ambient temperature is used to determine whether the vehicle is operating in low-temperature conditions (e.g., T < 0.05). 15℃).

[0175] 2) Determine the road surface type using cameras and radar.

[0176] For example, road surface images captured by an optical camera are combined with reflection signals obtained by millimeter-wave radar, and image recognition algorithms are used to analyze the differences in the distribution of laser reflection intensity of different materials (such as dry asphalt, wet asphalt, concrete, etc.) to determine the specific type of the current road surface.

[0177] For example, road surface types can include at least one of the following: dry asphalt, wet asphalt, concrete, dirt road, and water road.

[0178] 3) Obtain road surface texture features, water film or snow and ice coverage through cameras.

[0179] For example, an optical camera is used to capture images of the road surface ahead, extract the texture features of the road surface from the images, and determine whether the road surface is covered with water film, snow or ice to assess the slipperiness and adhesion conditions of the road surface.

[0180] 4) Estimate the real-time road adhesion coefficient based on road texture features, road type, and water film or snow cover.

[0181] For example, information such as road surface type, texture features, and water / snow cover is used as input and substituted into the tire-road adhesion Burckhardt model to calculate the real-time adhesion coefficient μ of the current road surface. The adhesion coefficient is used to quantify the maximum adhesion that the road surface can provide.

[0182] 5) Based on the vehicle's current location, obtain meteorological information within a radius of the current location and a first distance.

[0183] For example, the vehicle can receive weather warnings within a three-kilometer radius of its current location via a V2X communication module. These warnings can be used to predict whether there will be a warming trend in the next 30 minutes, leading to ice melting, increased rainfall, or other weather changes, thus allowing for timely adjustments to energy management strategies.

[0184] 6) Read battery management system data.

[0185] For example, battery data reported by the BMS can be read through the vehicle's CAN bus. The battery data may include battery status parameters such as average battery pack temperature, SOC, and internal resistance.

[0186] 7) Obtain the wheel speed of at least one wheel using a wheel speed sensor.

[0187] For example, wheel speed sensors on each wheel collect wheel speed data in real time. The wheel speed is used to determine whether the wheel speed difference exceeds the normal range (e.g., the wheel speed difference of the four wheels exceeds 5%~15%), so as to identify whether the wheels are at risk of slipping.

[0188] 8) Obtain the driver's operation request.

[0189] For example, the driver's driving intention can be determined by reading the accelerator pedal opening signal and its rate of change, and combining it with the brake pedal signal.

[0190] For example, after acquiring multi-source sensing information, it is necessary to align the multi-source sensing information in terms of time and space dimensions.

[0191] Time dimension alignment refers to unifying different sensor data collected at different sampling frequencies and trigger times to the same reference timestamp. Optionally, time dimension alignment includes: determining the reference timestamp; acquiring the original data of the preceding and following frames adjacent to the target timestamp; and calculating the pre-time alignment data of the target timestamp based on the original data of the preceding and following frames using interpolation.

[0192] Optionally, the camera exposure start time is used as the global master timestamp; the original data of the two frames before and after the target timestamp are obtained from other sensors (such as radar); assuming that the target moves at a constant speed in a straight line in a very short time, the position of the target at the target timestamp is calculated based on the position and velocity information of the frames before and after using linear interpolation as the data corresponding to the target timestamp.

[0193] For example, the camera exposes and images at 50ms, while the millimeter-wave radar outputs a point cloud frame at 40ms and 80ms. Using linear interpolation, the target position and velocity in the two radar frames at 40ms and 80ms are used to calculate the most likely precise position of the radar target at 50ms, which is then aligned and fused with the road image from the camera.

[0194] Spatial dimension alignment refers to projecting radar and camera data from different installation locations and coordinate systems into a unified vehicle coordinate system, enabling a one-to-one correspondence between the detection data from multiple sensors in three-dimensional space. Spatial dimension alignment includes: establishing a vehicle coordinate system; and projecting radar and camera data into the vehicle coordinate system.

[0195] Optionally, a vehicle coordinate system is established with the rear axle center (or centroid) of the vehicle as the origin, the X-axis pointing forward, the Y-axis pointing left, and the Z-axis pointing upward. For radar data, its polar coordinates (distance r, azimuth α) are first solved into local radar Cartesian coordinates, and then the rotation matrix R and translation vector t obtained through joint calibration are projected onto the vehicle coordinate system. For camera data, the pixel coordinates (u,v) are back-projected into three-dimensional points in the camera coordinate system using the pinhole camera inverse projection model, combined with depth estimation, and then mapped to the vehicle coordinate system through the same extrinsic parameter matrix (rotation matrix R and translation vector t).

[0196] For example, there is a stationary vehicle 20 meters directly in front of the main vehicle. The radar measures the target distance r = 20.2m and the azimuth angle α = 0.5° (leftward) in polar coordinates. After trigonometric function calculation, the radar's local coordinates are obtained. Then, the radar's extrinsic parameter matrix (R = identity matrix, t = [+2.0, 0, 0]) is projected onto the vehicle's coordinate system, giving the target's position in the vehicle's coordinate system as approximately (22.2m, 0.18m, 0m). Simultaneously, the camera detects the bounding box of the same vehicle in pixel coordinates. Using depth estimation, it obtains a distance of 20m. After camera inverse projection and extrinsic parameter matrix projection, the target's position in the vehicle's coordinate system is given as approximately (21.5m, 0.15m, 0m). Since the two sensing sources have the same spatial location (approximately 22m in front and approximately 0.15~0.18m to the left), the system can determine that they are the same target, achieving spatial fusion of radar and camera data.

[0197] Step 520: Identify the vehicle's current operating condition based on the vehicle's multi-source perception information.

[0198] For example, multi-source sensing information is input into a pre-trained working condition recognition model, and temperature, adhesion coefficient and vehicle state are fuzzified and mapped to fuzzy sets; the fuzzy sets include the fuzzy types corresponding to the multi-source sensing information respectively.

[0199] Fuzzification refers to mapping the precise value of each parameter to a corresponding fuzzy set using a predefined membership function, assigning it a membership degree (ranging from 0 to 1) to each fuzzy type. For example, temperature is mapped to a low / medium / high fuzzy set; adhesion coefficient is mapped to an extremely low / low / medium / high fuzzy set; and vehicle status is mapped to fuzzy sets such as start / cruise / braking.

[0200] For example, in multi-source sensing information, the ambient temperature is -15℃. The temperature fuzzy set includes [-∞, +0.5℃] as low, [-0.5℃, 26.5℃] as medium, and [25.5℃, +∞] as high. Then, the membership degree of the ambient temperature -15℃ is 1 in the low fuzzy type and 0 in the medium and high fuzzy types.

[0201] It should be noted that fuzzy sets, fuzzy types, and membership functions can be set as needed; the above are just examples.

[0202] Then, the fuzzy set is matched with the antecedent conditions of the candidate rules in the rule base, the membership degree between the fuzzy type in the fuzzy set and the antecedent conditions is calculated, and the activation strength of the candidate rule is calculated based on the membership degree; the activation strength is used to indicate the degree of matching between the fuzzy set and the antecedent conditions.

[0203] After fuzzifying the multi-source sensing information, a pre-set rule base is read. The rule base stores the correspondence between antecedent conditions and candidate conditions (a candidate rule includes an antecedent condition and a candidate condition). For each candidate rule, the antecedent condition is matched with the fuzzy set of multi-source sensing information to obtain the membership degree value of the fuzzy set on the antecedent condition. Then, the activation strength of the candidate rule is calculated by taking the minimum membership degree or multiplying the membership degrees.

[0204] For example, suppose the multi-source sensing information includes T= Given a temperature of 18℃, μ=0.25, and the vehicle in a starting state, the first candidate rule in the rule base is: IF Low Temperature AND Extremely Low Adhesion Coefficient AND Vehicle Starting THEEN Extreme Cold Low Adhesion Mode; the second candidate rule is: IF Medium Temperature AND Medium Adhesion Coefficient AND Vehicle Cruise THEEN Normal Temperature Cruise Mode. Here, Low Temperature AND Extremely Low Adhesion Coefficient AND Vehicle Starting is the antecedent condition of the first candidate rule, and Extreme Cold Low Adhesion Mode is the candidate condition for the first candidate rule. Similarly, Medium Temperature AND Medium Adhesion Coefficient AND Vehicle Cruise is the antecedent condition of the second candidate rule, and Normal Temperature Cruise Mode is the candidate condition for the second candidate rule. Therefore, by matching the multi-source sensing information with the antecedent conditions of each candidate rule and calculating the membership degree, the membership degree of the multi-source sensing information in the first candidate rule is 1 for Low Temperature, 1 for Extremely Low Adhesion Coefficient, and 1 for Vehicle Starting; thus, the activation strength of the first candidate rule (the minimum of the three membership degrees) is 1 (perfect match). In the second candidate rule, the membership degree of the multi-source perception information is 0 for temperature, 0 for adhesion coefficient, and 0 for vehicle cruise. Therefore, the activation intensity of the second candidate rule is 0 (complete mismatch).

[0205] Finally, the target rule is determined based on the activation intensity of the candidate rules, and the working condition corresponding to the target rule is output as the current working condition.

[0206] After calculating the activation intensity of all candidate rules in the rule base based on multi-source sensing information, the final target rule is determined through a competition mechanism. For example, the candidate rule with the highest activation intensity is selected as the target rule, and the remaining rules are suppressed. Then, the corresponding working condition in the consequent (THEN part, i.e., candidate working condition) of the target rule is directly output as the current recognition result.

[0207] In one optional embodiment, the controller may also activate the corresponding management sub-policy based on the current operating conditions.

[0208] In the current complex operating conditions of low temperature and low adhesion, the management sub-strategies include: limiting the rise rate of the motor peak torque request to within a preset safety envelope; reducing the kinetic energy recovery intensity; sending instructions to the chassis domain controller to increase the damping and controlling the ESP to actively establish the brake master cylinder pre-pressure.

[0209] For example, the rise rate of the motor's peak torque request is limited to a preset safety envelope (e.g., <50 Nm / 100 ms). When the driver presses the accelerator pedal hard, the motor will not instantaneously release torque exceeding the static friction limit under extremely low adhesion coefficients, thus preventing the wheels from slipping and losing control.

[0210] For example, reducing the intensity of kinetic energy recovery from the standard mode 0.3g reduced to At 0.1g, hydraulic mechanical braking is the primary method of braking. On low-friction surfaces, excessive regenerative braking torque on a single axle can easily cause rear wheel lock-up or yaw instability. Reducing the regenerative braking intensity can distribute the braking force more evenly to all four wheels, ensuring longitudinal and lateral stability.

[0211] For example, a command is sent to the chassis domain controller to increase the damper damping to suppress vehicle pitch, while simultaneously instructing the ESP to actively build up pre-charge pressure in the master cylinder. Pre-charge pressure can shorten the braking response time by approximately 200ms, allowing braking force to be transmitted to the wheel ends more quickly when the driver needs to brake urgently, thus reducing braking distance on roads with extremely low traction.

[0212] In one alternative embodiment, the controller also extrapolates the trajectory of terminal voltage changes over future periods based on the vehicle's current real-time current, battery state of charge (SOC), and internal resistance.

[0213] If the predicted terminal voltage in the terminal voltage change trajectory reaches the undervoltage lower limit, a power cap warning is sent to the VCU; the VCU is used to perform forward filtering on the accelerator pedal opening in response to the power cap warning.

[0214] For example, the controller collects real-time data on the vehicle's current current, battery SOC, and internal resistance. Combined with a battery equivalent circuit model, it predicts the terminal voltage change trajectory over a continuous period. When the predicted terminal voltage is about to reach the undervoltage protection lower limit, the controller sends a power capping warning signal to the VCU before the voltage actually drops. Upon receiving this warning, the VCU does not need to wait until the voltage actually drops to the undervoltage lower limit before passively limiting torque. Instead, it performs forward filtering on the accelerator pedal opening request, actively and smoothly reducing the slope of the driver's torque request, allowing the motor output power to smoothly and early enter the capping state. This avoids sudden power interruption caused by a sudden voltage drop triggering battery hard protection, and also prevents the driver from repeatedly triggering "power interruption-recovery" oscillations due to rapid acceleration under low SOC conditions. This maintains the smoothness of vehicle power output and driving experience while ensuring battery safety.

[0215] In an alternative embodiment, the controller may also invoke a neural network model or a magic formula tire model to predict the vehicle’s expected slip ratio based on the current expected torque.

[0216] When the expected slip ratio reaches or exceeds the optimal slip ratio, a first-order inertial low-pass filter is superimposed on the torque request front end; wherein, the time constant of the first-order inertial low-pass filter is dynamically adjusted according to the expected slip ratio.

[0217] For example, the controller invokes a pre-trained neural network model or the Pacejka Magic Formula tire model to predict the expected slip ratio change trend of the wheel within the next 1-2 seconds based on the currently applied expected torque. When the predicted slip ratio is about to reach or exceed the optimal slip ratio (approximately 15%-20%, i.e., the critical point where the tire-road adhesion coefficient reaches its peak), the system determines that without intervention, the tire will enter a non-linear runaway region of deep slippage. To address this proactive judgment, the controller superimposes a first-order inertial low-pass filter at the torque request front end, "smoothing" the originally steep torque change curve into a gently descent slope. The time constant τ of the first-order inertial low-pass filter is not a fixed value but dynamically adjusted according to the expected slip ratio. The closer the slip ratio is to or beyond the optimal value, the larger τ is, resulting in a stronger filtering effect and a smoother torque change. This avoids the wheel exceeding its adhesion limit while controlling the longitudinal impact of the vehicle body within a comfortable threshold (<2m / s²). 3 )Inside.

[0218] Step 531: Adjust the objective function of energy management according to the current operating conditions, and adjust the constraints of vehicle control according to the current operating conditions. The objective function includes the weighted sum of energy consumption cost indicators, battery temperature maintenance indicators, and driving safety indicators; the constraints include thermal management strategies and / or torque distribution and anti-slip control.

[0219] For example, the energy consumption cost index can be adjusted according to the current operating conditions to determine the corresponding energy consumption economic weight.

[0220] And / or, adjust the battery temperature maintenance weight corresponding to the battery temperature maintenance index according to the current operating conditions.

[0221] And / or, adjust the driving safety weights of driving safety indicators according to the current operating conditions.

[0222] For example, the initial value of the economy weight is the first value; the initial value of the battery temperature maintenance weight is the second value; and the initial value of the driving safety weight is the third value. The first value is greater than the second value, and the first value is greater than the third value.

[0223] Under the current complex working conditions of low temperature and low adhesion, the energy consumption cost index corresponding to the energy consumption economy weight is reduced from the first value to the fourth value.

[0224] And / or, under the current working conditions of low temperature and low adhesion complex conditions, the weight of the battery temperature maintenance index is increased from the second value to the fifth value.

[0225] And / or, under the current working conditions of low temperature and low adhesion complex conditions, the driving safety weight corresponding to the driving safety index will be increased from the third value to the sixth value.

[0226] Among them, the fifth value is greater than the sixth value, and the sixth value is greater than the fourth value.

[0227] For example, under normal operating conditions, the weights are allocated as follows: energy consumption economy weight w1=0.6, battery temperature maintenance weight w2=0.2, and driving safety weight w3=0.2. At this time, the system prioritizes energy consumption economy as the main optimization objective, while battery temperature and driving safety are secondary constraints. However, when the system identifies complex operating conditions with low temperature and low adhesion, the weights are dynamically reconfigured to w1=0.1, w2=0.5, and w3=0.4. The energy consumption economy weight drops significantly from 0.6 to 0.1, no longer pursuing extreme power saving and allowing for appropriate sacrifice of energy consumption; the battery temperature maintenance weight increases to 0.5, forcing the system to prioritize ensuring the battery's thermal state and avoid reduced battery activity at low temperatures; and the driving safety weight increases to 0.4, making anti-skid and anti-loss-of-control important evaluation indicators.

[0228] For example, adjusting the thermal management strategy based on the current operating conditions; and / or adjusting torque distribution and anti-slip control based on the current operating conditions.

[0229] In the current complex operating conditions of low temperature and low adhesion, the thermal management strategy includes at least one of the following: 1) Turn off the battery cooling circuit.

[0230] For example, in extremely cold environments, the battery does not need to dissipate heat. Closing the cooling circuit can prevent cold air or low-temperature coolant from further carrying away battery heat and avoid the battery temperature from continuing to drop.

[0231] 2) Open the coolant series valve between the motor controller MCU and the battery pack.

[0232] For example, the cooling circuit of the MCU is connected in series with the cooling circuit of the battery pack. The waste heat generated by the motor is used to heat the coolant, and then the heat is transferred to the battery pack through the series circuit to realize waste heat recovery and assist the battery in heating up.

[0233] 3) When the motor is not working, the positive temperature coefficient PTC heater is instructed to heat the coolant at maximum power.

[0234] For example, when the motor is not working (no residual heat is available), the PTC heater is controlled to directly intervene at maximum power (limited by the low-voltage battery capacity) to actively heat the coolant and transfer the heat to the battery pack through a series circuit, thus forcibly warming the battery.

[0235] 4) Adjust the battery temperature rise target curve to raise the battery temperature to a first temperature within a first time period; the first temperature is determined based on the battery charging power limit temperature.

[0236] For example, a target curve for rapid battery temperature rise is set, requiring the cell temperature to rise above 0°C (first temperature) within 300 seconds (first duration) to remove the charging power limit and reduce the battery's internal resistance. The first temperature is determined based on the battery's charging power limit temperature; for example, 0°C is the temperature threshold for removing the battery's power limit.

[0237] Torque distribution and anti-slip control include at least one of the following: 1) Under the current working conditions of low temperature and low adhesion complex conditions, perform first-order low-pass filtering on the requested torque corresponding to the accelerator pedal request; and increase the system response time coefficient of the accelerator pedal: the time coefficient is used to control the speed at which the system responds to the accelerator pedal request.

[0238] For example, based on the torque requested by the driver, a first-order low-pass filter is introduced, with its time constant τ increased from the conventional 0.1s to 0.5s. The time constant is used to control the speed at which the system responds to the accelerator pedal request. The larger τ is, the slower the system response and the smoother the torque change, thereby eliminating sudden torque changes and preventing instantaneous impacts from exceeding the road adhesion limit.

[0239] 2) Under the current working conditions of low temperature and low adhesion, the maximum adhesion of the wheel is dynamically calculated based on the real-time estimated road adhesion coefficient; the output torque of the motor is limited based on the maximum adhesion and safety factor.

[0240] For example, the maximum adhesion force F of each wheel is dynamically calculated based on the real-time estimated road adhesion coefficient μ. max =μ·F z (F) z (for vertical loads), and limit the motor output torque to T. limit =F max ·r wheel / i gear ·K safe K safe To ensure a safety factor (0.8), the motor output torque is always kept below the adhesion limit provided by the road surface to prevent slippage.

[0241] 3) When the current working condition is a complex working condition of low temperature and low adhesion, and the slip ratio of the first wheel is detected to be higher than the slip ratio threshold, reduce the output torque of the first motor and increase the output torque of the second electrode.

[0242] For example, when the slip ratio of a wheel on one side (the first side) is detected to exceed the threshold (5%), the system immediately implements millisecond-level torque reduction on the motor on that side, and at the same time transfers the torque vector to the side with better adhesion (the second side). By using differential torque control, the system maintains the vehicle's yaw stability and driving force, ensuring that the vehicle does not deviate from its driving trajectory even when the adhesion coefficient on one side of the road is extremely low.

[0243] Step 541: Input the objective function and constraints into the MPC, and generate control commands based on the current state of the vehicle and the operation request; the predictive controller is used to generate control sequences in the rolling time domain, and the control sequences include at least one control command.

[0244] For example, the energy consumption cost index of the objective function includes a vehicle speed tracking deviation term, and the battery temperature maintenance index of the objective function includes a thermal coupling cost term; the driving safety index of the objective function includes a desired acceleration deviation term and a motor torque increment penalty term; the objective function is used to evaluate the quality of control commands; the vehicle speed tracking deviation term is used to penalize the deviation between the actual vehicle speed and the target vehicle speed; the desired acceleration deviation term is used to penalize the deviation between the actual acceleration and the desired acceleration; the motor torque increment penalty term is used to penalize the change in motor torque within adjacent control cycles; and the thermal coupling cost term is used to penalize the deviation between the actual power and the target power of the PTC heater.

[0245] Based on the constraints and with the objective function as the minimum, the online quadratic programming solver of MPC iteratively solves the problem in the rolling time domain based on the vehicle's current state and operation requests, and outputs the control sequence.

[0246] The control commands include at least one of the following: target motor torque, PTC heating power, and ESP intervention threshold.

[0247] Optionally, the objective function (J) is input into the MPC along with the hard constraint boundaries. This objective function consists of four terms: a speed tracking deviation term (k1·||Δv||²) to penalize the deviation between the actual vehicle speed and the target vehicle speed, ensuring the vehicle follows the driver's intentions; a desired acceleration deviation term (k2·||Δa||²) to penalize the deviation between the actual acceleration and the desired acceleration, suppressing longitudinal impact; and a motor torque increment penalty term (k3·||ΔT). motor ||²) is used to penalize the change in motor torque within adjacent control cycles, preventing slippage caused by sudden torque changes; thermal coupling cost term (k4·(P) PTC P target)²) is used to penalize the deviation between the actual power and the target power of the PTC heater, ensuring that battery thermal management is performed as needed.

[0248] With the goal of minimizing J, MPC predicts N in the time domain. p Within a rolling time window of 20 steps and a control step size Δt = 100ms (i.e., a total prediction time of 2.0 seconds), the online QP solver is invoked to iterate and solve for the optimal control sequence cycle by cycle based on the vehicle's current state (wheel speed, SOC, terminal voltage, adhesion coefficient, etc.) and operation requests.

[0249] In each control cycle, MPC only issues the first control quantity in the sequence, namely the target torque of the motor after gradient limiting, the PTC heating power, and the ESP intervention threshold, and feeds back the actual execution result to the prediction model input in the next cycle to cover the prediction error of the previous moment.

[0250] Step 542: Collect the vehicle's center of gravity speed using the chassis domain controller at the first sampling frequency, and calculate the longitudinal slip ratio.

[0251] For example, the chassis domain controller acquires the wheel speed measured by the wheel speed sensor and the vehicle center-of-gravity speed calculated by the INS in real time at a first sampling frequency of 1 kHz (every 1 ms), and calculates the real-time longitudinal slip ratio using the wheel speed and vehicle center-of-gravity speed. The longitudinal slip ratio is used to characterize the degree of slippage of the wheel relative to the road surface during driving or braking, and can be used to determine whether the tire is at its adhesion limit.

[0252] Step 543: When the longitudinal slip ratio is within the normal range, execute the control command output by MPC.

[0253] The maximum value in the normal range is less than the minimum value in the subcritical range. That is, under normal longitudinal slip ratio, the vehicle's actuators execute the control commands generated by the MPC normally.

[0254] Step 544: When the longitudinal slip ratio is in the subcritical range, perform soft correction on the objective function and / or constraints in MPC.

[0255] When the longitudinal slip ratio is in the subcritical range of [5%, 15%], it is determined that the tire has entered the edge of the nonlinear adhesion region but is still within a controllable and stable slip range. At this time, MPC does not intervene aggressively, but instead initiates online weight self-tuning soft correction. The boundary constraints of the QP solver are tightened, compressing the maximum allowable torque change in each control cycle from the usual 200Nm to 50Nm, which is equivalent to applying a first-order low-pass filter to the accelerator pedal request, slowing down the power release rate. The objective function weights are temporarily adjusted, increasing the penalty for acceleration stability and appropriately sacrificing some vehicle speed tracking accuracy, so that the optimizer prioritizes suppressing impacts rather than pursuing vehicle speed following. At the same time, the rolling time domain is contracted. If the slip ratio shows an upward trend, the prediction model inside MPC will adjust the virtual maximum ground friction constraint downward in advance, so that the torque command in the control sequence of the next few steps will automatically have attenuation feedforward, achieving smooth domestication by pre-reducing before exceeding the limit.

[0256] For example, soft corrections include at least one of the following: 1) Reduce the torque increment step size.

[0257] The torque increment step size is the maximum change in motor torque relative to the output value of the previous cycle that the QP solver allows within each control cycle (i.e., the limit of the rise or fall).

[0258] 2) Adjust the weight of at least one of the following in the objective function: vehicle speed tracking deviation, thermal coupling cost, expected acceleration deviation, and motor torque increment penalty.

[0259] For example, increase the weight of the expected acceleration deviation term and decrease the weight of the vehicle speed tracking deviation term.

[0260] 3) Adjust the boundaries of the constraints.

[0261] For example, when the longitudinal slip ratio is increasing, the maximum friction force of the virtual ground is reduced in advance; the maximum friction force of the virtual ground is used to constrain the maximum value of the longitudinal force of the wheel.

[0262] Step 545: If the longitudinal slip ratio is higher than the slip ratio threshold, do not execute the control command output by MPC, and execute the emergency intervention command through HSM; the slip ratio threshold is not lower than the maximum value of the subcritical range.

[0263] For example, emergency intervention instructions include at least one of the following: 1) Send a PWM (Pulse Width Modulation) blocking signal to the gate power amplifier of the inverter to reduce the three-phase drive current to a safe shutdown state and force the motor output torque to return to zero.

[0264] When the slip ratio exceeds the critical threshold of 15%, the tire has completely exceeded the adhesion ellipse limit and entered a deep slip / free-spinning state, at which point the mathematical basis of MPC based on the nominal model no longer holds. At this point, the HSM immediately bypasses the MPC software output and directly sends a PWM blocking signal to the inverter's Gate Driver (gate-level power amplifier). The PWM blocking signal bypasses the CAN bus delay and reduces the three-phase drive current to a safe shutdown state within ≤10ms (usually only 2~4ms), forcing the motor output torque to zero, completely cutting off the power source, and preventing the slip ratio from diverging to 100% free-spinning due to continuous drive.

[0265] 2) The ESP active pressure building unit applies high-frequency pulse braking pressure to the target wheel that is experiencing longitudinal slippage.

[0266] While cutting off power, the ESP active pressure-building unit responds immediately, applying high-frequency pulsating braking pressure (i.e., intermittent braking) at a frequency of 15~20Hz to the target wheel (or multiple wheels) that has slipped. The high-frequency pulsating braking pressure fluctuations can effectively break the static friction stickiness between the tire and the ice surface, and in conjunction with the dynamic recovery of the ground adhesion coefficient, quickly pull the wheel speed back to near the reference vehicle speed, thereby assisting the vehicle to regain traction.

[0267] 3) If the longitudinal slip ratio is not lower than the safety threshold, control of the MPC will not be restored.

[0268] For example, the HSM will only release control and lift torque lockout when the slip ratio is forcibly lowered and remains below the 5% safety threshold for three consecutive control cycles (i.e., 300ms). Before this, as long as the slip ratio does not fall below the safety threshold, the HSM will continue to maintain control and not relinquish it to the MPC, thus avoiding control chatter caused by frequent switching at threshold boundaries.

[0269] 4) If the longitudinal slip ratio is below the safety threshold for N consecutive control cycles, release the control of the HSM and restore the control of the MPC. Based on the ramp function, the motor torque will be increased from 0 to the output torque calculated by the MPC within a preset time. N is a positive integer.

[0270] For example, when the slip ratio has been below the safety threshold of 5% for three consecutive control cycles, the HSM releases the torque lock and restores MPC control. However, the system does not instantly switch back to the MPC's request command. Instead, it uses a ramp function to smoothly ramp up the motor torque from 0 to the current MPC-calculated request value within 200ms, achieving a seamless transition between hard backoff and soft control, and avoiding shocks and secondary slippage caused by sudden torque recovery.

[0271] Step 546: Monitor the execution results of the control commands; the execution results include at least one of the actual wheel speed, the actual response torque of the motor, and the hydraulic cylinder pressure; the execution results are used as the current state of the vehicle and fed back to the MPC.

[0272] In summary, the method provided in this application acquires multi-source sensing information and identifies the vehicle's current operating condition in real time based on this information, accurately determining the vehicle's actual driving environment. Based on the vehicle's current operating condition, the objective function and constraints of energy management are adjusted, allowing the parameters of the energy management strategy to dynamically change with the vehicle's actual driving state. This enables automatic adaptation to the optimal control strategy even under non-standard operating conditions such as extreme cold, low traction, and congestion, improving the rationality of energy regulation and its generalization ability. Furthermore, the objective function can balance the three major objectives of energy management: economy, battery temperature, and driving safety. By reconstructing the weights of each objective based on the current operating condition and coordinating their priorities, it avoids using energy-saving strategies under normal operating conditions when the battery temperature is too high or the road surface traction is extremely low, effectively resolving the conflict and coupling problem between multiple objectives. Through optimization of the energy management strategy and precise control by intelligent algorithms, comprehensive energy consumption can be reduced by 5% to 10% under complex and variable road conditions (such as mountain climbing and urban congestion) and extreme climates (such as extreme cold and high temperatures).

[0273] The method provided in this application significantly increases the battery temperature maintenance weight from 0.2 to 0.5 under extremely cold and low-adsorption conditions, forcibly prioritizing the protection of the battery's thermal state. Through thermal management methods such as active heating with PTC heaters and waste heat recovery from electric drives, the cell temperature is rapidly increased, the charging power limitation is removed, the internal resistance is reduced, and the battery pack's service life is extended.

[0274] The method provided in this application does not perform aggressive intervention when the slip ratio is in the subcritical range of [5%, 15%]. Instead, it initiates online weighted self-tuning soft correction: compressing the maximum torque change amplitude per cycle from the conventional 200Nm to 50Nm; simultaneously increasing the acceleration stability penalty and appropriately sacrificing some vehicle speed tracking accuracy, so that the optimizer prioritizes suppressing impact. When the slip ratio is on the rise, the MPC prediction model will also adjust the virtual ground maximum friction constraint downward in advance, so that the torque command in the future control sequence will automatically have attenuation feedforward, actively reducing torque before the wheels reach deep slippage, thereby controlling the longitudinal impact of the vehicle body within the comfort threshold.

[0275] The method provided in this application embodiment allows the HSM to quickly bypass the MPC software output and send a PWM lockout signal to the inverter to force the torque to zero when the slip ratio exceeds the 15% threshold. This, combined with ESP high-frequency pulse braking, completely cuts off the power source, preventing vehicle loss of control. This hardware-level safety fallback mechanism provides assurance for driving safety under extreme physical limits.

[0276] The exemplary embodiments of this application have been described in detail above. It should be understood that those skilled in the art can make detailed modifications and variations based on the concept of this application without creative effort. Therefore, any technical solutions that can be obtained by those skilled in the art based on the concept of this application and through logical analysis, theoretical deduction, or limited experimentation on the basis of related technologies are intended to fall within the protection scope defined by this application.

[0277] Figure 7 This is a schematic diagram of the structure of an energy management device provided in an exemplary embodiment of this application. The device includes: The identification module 1002 is used to identify the current operating condition of the vehicle based on the vehicle's multi-source perception information; The adjustment module 1003 is used to adjust the objective function of energy management and the constraints of vehicle control according to the current operating conditions. The objective function is used to balance the economy of energy management, battery temperature and driving safety. The generation module 1004 is used to generate control instructions based on the objective function and the constraints, the control instructions being used to implement energy management that conforms to the objective function and the constraints.

[0278] In one optional embodiment, the objective function includes a weighted sum of energy consumption cost indicators, battery temperature maintenance indicators, and driving safety indicators; The adjustment module 1003 is used to adjust the energy consumption economic weight corresponding to the energy consumption cost index according to the current operating conditions. The adjustment module 1003 is used to adjust the battery temperature maintenance weight corresponding to the battery temperature maintenance index according to the current operating conditions. The adjustment module 1003 is used to adjust the driving safety weight of the driving safety index according to the current operating conditions.

[0279] In one optional embodiment, the initial value of the economy weight is a first value; the initial value of the battery temperature maintenance weight is a second value; and the initial value of the driving safety weight is a third value, wherein the first value is greater than the second value and the first value is greater than the third value. The adjustment module 1003 is configured to perform at least one of the following: When the current working condition is a complex working condition with low temperature and low adhesion, the energy consumption cost index corresponding to the energy consumption economy weight is reduced from the first value to the fourth value. When the current working condition is a complex working condition of low temperature and low adhesion, the weight of the battery temperature maintenance index is increased from the second value to the fifth value. When the current working condition is a complex working condition with low temperature and low adhesion, the driving safety weight corresponding to the driving safety index is increased from the third value to the sixth value. The fifth value is greater than the sixth value, and the sixth value is greater than the fourth value.

[0280] In an optional embodiment, the adjustment module 1003 is configured to adjust the thermal management strategy according to the current operating conditions; and / or, Adjust torque distribution and anti-slip control according to the current operating conditions.

[0281] In an optional embodiment, the adjustment module 1003 is configured to execute at least one of the following thermal management strategies when the current operating condition is a complex condition of low temperature and low adhesion: Shut down the battery cooling circuit; Open the coolant series valve between the motor controller MCU and the battery pack; When the motor is not working, the positive temperature coefficient PTC heater is instructed to heat the coolant at maximum power. The target curve for battery temperature rise is adjusted to raise the battery temperature to a first temperature within a first duration; the first temperature is determined based on the limiting temperature of the battery charging power.

[0282] In an optional embodiment, the adjustment module 1003 is configured to perform a first-order low-pass filter on the requested torque corresponding to the accelerator pedal request when the current operating condition is a complex condition of low temperature and low adhesion; and to increase the system response time coefficient of the accelerator pedal: the time coefficient is used to control the speed at which the system responds to the accelerator pedal request; and / or, Under the current operating condition of low temperature and low adhesion, the maximum adhesion of the wheels is dynamically calculated based on the real-time estimated road adhesion coefficient; the output torque of the motor is limited based on the maximum adhesion and a safety factor; and / or, When the current working condition is a complex condition of low temperature and low adhesion, and the slip ratio of the first wheel is detected to be higher than the slip ratio threshold, the output torque of the first motor is reduced and the output torque of the second electrode is increased.

[0283] In an optional embodiment, the generation module 1004 is configured to input the objective function and the constraints into the model predictive controller (MPC) and generate the control instructions based on the current state of the vehicle and the operation request; the predictive controller is configured to generate a control sequence in the rolling time domain, the control sequence including at least one control instruction.

[0284] In an optional embodiment, the energy consumption cost index of the objective function includes a vehicle speed tracking deviation term, and the battery temperature maintenance index of the objective function includes a thermal coupling cost term; the driving safety index of the objective function includes a desired acceleration deviation term and a motor torque increment penalty term; the objective function is used to evaluate the quality of control commands; the vehicle speed tracking deviation term is used to penalize the deviation between the actual vehicle speed and the target vehicle speed; the desired acceleration deviation term is used to penalize the deviation between the actual acceleration and the desired acceleration; the motor torque increment penalty term is used to penalize the change in motor torque within adjacent control cycles; and the thermal coupling cost term is used to penalize the deviation between the actual power and the target power of the PTC heater. The generation module 1004 is used to iteratively solve the problem in the rolling time domain based on the current state of the vehicle and the operation request, according to the constraints and with the objective function as the goal, through the online quadratic programming solver of the MPC, and output the control sequence.

[0285] In one alternative embodiment, the control commands include at least one of the following: target motor torque, PTC heating power, and ESP intervention threshold.

[0286] In an optional embodiment, the device further includes: The correction module 1005 is used to collect the vehicle center of gravity speed according to the first sampling frequency through the chassis domain controller and calculate the longitudinal slip ratio; The correction module 1005 is used to perform soft correction on the objective function and / or the constraint conditions in the MPC when the longitudinal slip ratio is in the subcritical range.

[0287] In one alternative embodiment, the correction module 1005 is configured to perform at least one of the following: Reduce the torque increment step size; Adjust the weight of at least one of the following in the objective function: vehicle speed tracking deviation, thermal coupling cost, expected acceleration deviation, and motor torque increment penalty. Adjust the boundaries of the constraints.

[0288] In an optional embodiment, the correction module 1005 is used to increase the weight of the desired acceleration deviation term and decrease the weight of the vehicle speed tracking deviation term.

[0289] In an optional embodiment, the correction module 1005 is used to pre-reduce the maximum friction force of the virtual ground when the longitudinal slip ratio is increasing; the maximum friction force of the virtual ground is used to constrain the maximum value of the longitudinal force of the wheel.

[0290] In an optional embodiment, the device further includes: The correction module 1005 is used to collect the vehicle center speed according to the first sampling frequency through the chassis domain controller and calculate the longitudinal slip ratio. The correction module 1005 is used to, when the longitudinal slip ratio is higher than the slip ratio threshold, not execute the control command output by the MPC, but execute an emergency intervention command through the hardware safety monitoring unit HSM; the slip ratio threshold is not lower than the maximum value of the subcritical range.

[0291] In one alternative embodiment, the correction module 1005 is configured to perform at least one of the following: Send a pulse width modulation (PWM) block signal to the gate power amplifier of the inverter to reduce the three-phase drive current to a safe shutdown state and force the motor output torque to zero. The ESP active pressure build-up unit applies high-frequency pulse braking pressure to the target wheel that is experiencing longitudinal slippage; If the longitudinal slip ratio is not lower than the safety threshold, control of the MPC will not be restored; If the longitudinal slip ratio is lower than the safety threshold for N consecutive control cycles, the control of the HSM is released and the control of the MPC is restored. Based on the ramp function, the motor torque is increased from 0 to the output torque calculated by the MPC within a preset time.

[0292] In an optional embodiment, the device further includes: The feedback module 1006 is used to monitor the execution result of the control command; the execution result includes at least one of the actual wheel speed, the actual response torque of the motor, and the hydraulic cylinder pressure. The feedback module 1006 is used to feed the execution result as the current state of the vehicle back to the MPC.

[0293] In an optional embodiment, the device further includes: Module 1001 is configured to perform at least one of the following: Use an external temperature sensor to obtain the ambient temperature; The road surface type is determined using cameras and radar; The camera captures road surface texture features, water film, or snow and ice coverage. The real-time road adhesion coefficient is estimated based on the road texture features, the road type, and the water film or snow cover. Based on the current location of the vehicle, obtain meteorological information within a radius of a first distance centered on the current location; Read data from the battery management system; The wheel speed of at least one wheel is obtained using a wheel speed sensor; Obtain the driver's operation request.

[0294] In an optional embodiment, the acquisition module 1001 is used to align the multi-source sensing information in terms of time and space dimensions. The time dimension alignment includes: determining a reference timestamp; obtaining the original data of the preceding and following frames adjacent to the target timestamp; and calculating the pre-time alignment data of the target timestamp based on the original data of the preceding and following frames using an interpolation method. The spatial dimension alignment includes: establishing a vehicle coordinate system; and projecting radar data and camera data into the vehicle coordinate system.

[0295] In an optional embodiment, the recognition module 1002 is used to input the multi-source sensing information into a pre-trained working condition recognition model, and to fuzzify the temperature, adhesion coefficient and vehicle state, mapping them into a fuzzy set; the fuzzy set includes the fuzzy types corresponding to the multi-source sensing information respectively; The recognition module 1002 is used to match the fuzzy set with the antecedent conditions of candidate rules in the rule base, calculate the membership degree between the fuzzy type in the fuzzy set and the antecedent conditions, and calculate the activation intensity of the candidate rule based on the membership degree; the activation intensity is used to indicate the degree of matching between the fuzzy set and the antecedent conditions. The identification module 1002 is used to determine the target rule based on the activation intensity of the candidate rule, and output the working condition corresponding to the target rule as the current working condition.

[0296] In an optional embodiment, the identification module 1002 is used to activate the corresponding management sub-policy according to the current working condition; In the case of the current working condition being a complex condition of low temperature and low adhesion, the management sub-strategy includes: limiting the rise rate of the motor peak torque request to within a preset safety envelope; reducing the kinetic energy recovery intensity; sending a command to the chassis domain controller to increase the damping and controlling the ESP to actively establish the brake master cylinder pre-pressure.

[0297] In an optional embodiment, the identification module 1002 is used to extrapolate the trajectory of terminal voltage changes in the future period based on the vehicle's current real-time current, battery state of charge (SOC), and internal resistance. The identification module 1002 is used to send a power cap warning to the vehicle controller (VCU) when the predicted terminal voltage in the terminal voltage change trajectory reaches the undervoltage lower limit; the VCU is used to perform forward filtering on the accelerator pedal opening in response to the power cap warning.

[0298] In an optional embodiment, the identification module 1002 is used to invoke a neural network model or a magic formula tire model to predict the expected slip ratio of the vehicle based on the current expected torque. The identification module 1002 is used to superimpose a first-order inertial low-pass filter on the torque request front end when the expected slip ratio reaches or exceeds the optimal slip ratio; wherein the time constant of the first-order inertial low-pass filter is dynamically adjusted according to the expected slip ratio.

[0299] It should be noted that the energy management device provided in the above embodiments is only an example of the division of the above functional modules. In practical applications, the above functions can be assigned to different functional modules as needed, that is, the internal structure of the device can be divided into different functional modules to complete all or part of the functions described above. In addition, the energy management device and the energy management method embodiments provided in the above embodiments belong to the same concept, and the specific implementation process can be found in the method embodiments, which will not be repeated here.

[0300] Figure 8 This is a schematic diagram of the structure of a vehicle-mounted terminal according to an embodiment of this application.

[0301] Typically, the vehicle terminal 1100 includes: a main control module 1101, a CAN interface 1102, a hard-wired input interface 1103, and a hard-wired output interface 1104. The main control module 1101 is connected to the CAN interface 1102, the hard-wired input interface 1103, and the hard-wired output interface 1104, respectively.

[0302] The main control module 1101 typically includes a processor and memory. The processor may include one or more processing cores, such as an 11-core processor or an 8-core processor. The processor can be implemented using at least one hardware form of DSP (Digital Signal Processing), FPGA (Field-Programmable Gate Array), or PLA (Programmable Logic Array). The processor may also include a main processor and a coprocessor. The main processor, also known as a CPU (Central Processing Unit), is used to process data in the wake-up state; the coprocessor is a low-power processor used to process data in the standby state. In some embodiments, the processor may integrate a GPU (Graphics Processing Unit), which is responsible for rendering and drawing the content required to be displayed on the vehicle's display screen. In some embodiments, the processor may also include an AI (Artificial Intelligence) processor, which is used to handle computational operations related to machine learning. The memory may include one or more computer-readable storage media, which may be non-transitory. The memory may also include high-speed random access memory and non-volatile memory, such as one or more disk storage devices or flash memory devices. In some embodiments, a non-transitory computer-readable storage medium in the memory is used to store at least one piece of program code, which is executed by a processor to implement the energy management method provided in the method embodiments of this application.

[0303] The CAN interface 1102 may include a powertrain CAN interface, a motor CAN interface, and a diagnostic CAN interface. The powertrain CAN interface is used to communicate with the vehicle's powertrain module, the motor CAN interface is used to communicate with the vehicle's motor controller, and the diagnostic CAN interface is used to communicate with diagnostic equipment.

[0304] The hard-wired input interface 1103 is used to receive hard-wired control signals. The hard-wired output interface 1104 is used to send control commands to the vehicle's electronic control components, causing the vehicle's electronic control components to perform corresponding actions. The vehicle's electronic control components include a power management system, a motor controller, an on-board charger, and a body control system.

[0305] The main control module 1101 can communicate with the vehicle's powertrain module, motor controller, and diagnostic equipment via the CAN interface 1102, and generate control commands based on the hard-wired control signals received by the hard-wired input interface 1103, so as to send the control commands to the vehicle's electronic control components via the hard-wired output interface 1104.

[0306] Those skilled in the art will understand that Figure 8 The structure shown does not constitute a limitation on the vehicle terminal 1100, and may include more or fewer components than shown, or combine certain components, or use different component arrangements.

[0307] This application also provides a computer-readable storage medium storing at least one instruction, at least one program, code set, or instruction set. When the at least one instruction, at least one program, code set, or instruction set is loaded and executed by the processor of a computer device, the energy management method provided in the above-described method embodiments is implemented.

[0308] This application also provides a computer program product or computer program including computer instructions stored in a computer-readable storage medium. A processor of a computer device reads the computer instructions from the computer-readable storage medium and executes the computer instructions, causing the computer device to perform the energy management methods provided in the above-described method embodiments.

[0309] Those skilled in the art will understand that all or part of the steps of the above embodiments can be implemented by hardware or by a program instructing related hardware. The program can be stored in a computer-readable storage medium, such as a read-only memory, a disk, or an optical disk.

[0310] The above description is merely an optional embodiment of this application and is not intended to limit this application. Any modifications, equivalent switching, improvements, etc., made within the spirit and principles of this application should be included within the protection scope of this application.

Claims

1. An energy management method, characterized in that, The method includes: Based on the vehicle's multi-source perception information, the current operating condition of the vehicle is identified; The objective function of energy management and the constraints of vehicle control are adjusted according to the current operating conditions. The objective function is used to balance the economy of energy management, battery temperature and driving safety. Control commands are generated based on the objective function and the constraints, and these control commands are used to implement energy management that conforms to the objective function and the constraints.

2. The method according to claim 1, characterized in that, The objective function includes a weighted sum of energy consumption cost indicators, battery temperature maintenance indicators, and driving safety indicators; The objective function for adjusting energy management based on the current operating conditions includes at least one of the following: Adjust the energy consumption cost index corresponding to the energy consumption economic weight according to the current operating conditions; Adjust the battery temperature maintenance weight corresponding to the battery temperature maintenance index according to the current operating conditions. The driving safety weights of the driving safety indicators are adjusted according to the current operating conditions.

3. The method according to claim 2, characterized in that, The initial value of the economic efficiency weight is a first value; the initial value of the battery temperature maintenance weight is a second value; the initial value of the driving safety weight is a third value; the first value is greater than the second value; and the first value is greater than the third value. The objective function for adjusting energy management based on the current operating conditions includes at least one of the following: When the current working condition is a complex working condition with low temperature and low adhesion, the energy consumption cost index corresponding to the energy consumption economy weight is reduced from the first value to the fourth value. When the current working condition is a complex working condition of low temperature and low adhesion, the weight of the battery temperature maintenance index is increased from the second value to the fifth value. When the current working condition is a complex working condition with low temperature and low adhesion, the driving safety weight corresponding to the driving safety index is increased from the third value to the sixth value. The fifth value is greater than the sixth value, and the sixth value is greater than the fourth value.

4. The method according to any one of claims 1 to 3, characterized in that, The constraints for adjusting vehicle control include: Adjust the thermal management strategy according to the current operating conditions; and / or, Adjust torque distribution and anti-slip control according to the current operating conditions.

5. The method according to claim 4, characterized in that, The adjustment of the thermal management strategy according to the current operating conditions includes: When the current operating condition is a complex condition of low temperature and low adhesion, at least one of the following thermal management strategies shall be implemented: Shut down the battery cooling circuit; Open the coolant series valve between the motor controller MCU and the battery pack; When the motor is not working, the positive temperature coefficient PTC heater is instructed to heat the coolant at maximum power. The target curve for battery temperature rise is adjusted to raise the battery temperature to a first temperature within a first duration; the first temperature is determined based on the limiting temperature of the battery charging power.

6. The method according to claim 4, characterized in that, The adjustment of torque distribution and anti-slip control based on the current operating conditions includes: Under the current operating condition of low temperature and low adhesion complex conditions, a first-order low-pass filter is applied to the requested torque corresponding to the accelerator pedal request; and the system response time coefficient of the accelerator pedal is increased: the time coefficient is used to control the speed at which the system responds to the accelerator pedal request; and / or, Under the current operating condition of low temperature and low adhesion, the maximum adhesion of the wheels is dynamically calculated based on the real-time estimated road adhesion coefficient; the output torque of the motor is limited based on the maximum adhesion and a safety factor; and / or, When the current working condition is a complex condition of low temperature and low adhesion, and the slip ratio of the first wheel is detected to be higher than the slip ratio threshold, the output torque of the first motor is reduced and the output torque of the second electrode is increased.

7. The method according to any one of claims 1 to 3, characterized in that, The step of generating control instructions based on the objective function and the constraints includes: The objective function and the constraints are input into the model predictive controller (MPC), which generates the control commands based on the vehicle's current state and operation requests. The predictive controller generates a control sequence in the rolling time domain, and the control sequence includes at least one control command.

8. The method according to claim 7, characterized in that, The objective function's energy consumption cost index includes a vehicle speed tracking deviation term; the objective function's battery temperature maintenance index includes a thermal coupling cost term; the objective function's driving safety index includes a desired acceleration deviation term and a motor torque increment penalty term; the objective function is used to evaluate the quality of control commands; the vehicle speed tracking deviation term is used to penalize the deviation between the actual vehicle speed and the target vehicle speed; the desired acceleration deviation term is used to penalize the deviation between the actual acceleration and the desired acceleration; the motor torque increment penalty term is used to penalize the change in motor torque within adjacent control cycles; the thermal coupling cost term is used to penalize the deviation between the actual power and the target power of the PTC heater; The step of inputting the objective function and the constraints into the model predictive controller (MPC) to generate the control instructions includes: Based on the constraints and with the objective function as the goal, the online quadratic programming solver of the MPC iteratively solves the problem in the rolling time domain based on the current state of the vehicle and the operation request, and outputs a control sequence.

9. The method according to claim 7, characterized in that, The control commands include at least one of the following: target motor torque, PTC heating power, and ESP intervention threshold.

10. The method according to claim 7, characterized in that, The method further includes: The vehicle's center of gravity speed is collected by the chassis domain controller at the first sampling frequency, and the longitudinal slip ratio is calculated. When the longitudinal slip ratio is in the subcritical range, the objective function and / or the constraints in the MPC are softly modified.

11. The method according to claim 10, characterized in that, The soft modification of the objective function and / or constraints in the MPC includes at least one of the following: Reduce the torque increment step size; Adjust the weight of at least one of the following in the objective function: vehicle speed tracking deviation, thermal coupling cost, expected acceleration deviation, and motor torque increment penalty. Adjust the boundaries of the constraints.

12. The method according to claim 11, characterized in that, The weights corresponding to at least one of the following in the adjustment objective function: vehicle speed tracking deviation term, thermal coupling cost term, desired acceleration deviation term, and motor torque increment penalty term, include: Increase the weight of the expected acceleration deviation term and decrease the weight of the vehicle speed tracking deviation term.

13. The method according to claim 11, characterized in that, The adjustment of the boundary of the constraint conditions includes: When the longitudinal slip ratio is increasing, the maximum friction force of the virtual ground is reduced in advance; the maximum friction force of the virtual ground is used to constrain the maximum value of the longitudinal force of the wheel.

14. The method according to claim 7, characterized in that, The method further includes: The vehicle's center of gravity speed is collected by the chassis domain controller at the first sampling frequency, and the longitudinal slip ratio is calculated. If the longitudinal slip ratio is higher than the slip ratio threshold, the control command output by the MPC is not executed, and an emergency intervention command is executed through the hardware safety monitoring unit (HSM); the slip ratio threshold is not lower than the maximum value of the subcritical range.

15. The method according to claim 14, characterized in that, The execution of emergency intervention instructions includes at least one of the following: Send a pulse width modulation (PWM) block signal to the gate power amplifier of the inverter to reduce the three-phase drive current to a safe shutdown state and force the motor output torque to zero. The ESP active pressure build-up unit applies high-frequency pulse braking pressure to the target wheel that is experiencing longitudinal slippage; If the longitudinal slip ratio is not lower than the safety threshold, control of the MPC will not be restored; If the longitudinal slip ratio is lower than the safety threshold for N consecutive control cycles, the control of the HSM is released and the control of the MPC is restored. Based on the ramp function, the motor torque is increased from 0 to the output torque calculated by the MPC within a preset time period, where N is a positive integer.

16. The method according to claim 7, characterized in that, The method further includes: The execution results of monitoring and control commands include at least one of the following: actual wheel speed, actual motor response torque, and hydraulic cylinder pressure. The execution result is used as the current state of the vehicle and fed back to the MPC.

17. The method according to any one of claims 1 to 3, characterized in that, The method further includes acquiring the multi-source sensing information through at least one of the following methods: Use an external temperature sensor to obtain the ambient temperature; The road surface type is determined using cameras and radar; The camera captures road surface texture features, water film, or snow and ice coverage. The real-time road adhesion coefficient is estimated based on the road texture features, the road type, and the water film or snow cover. Based on the current location of the vehicle, obtain meteorological information within a radius of a first distance centered on the current location; Read data from the battery management system; The wheel speed of at least one wheel is obtained using a wheel speed sensor; Obtain the driver's operation request.

18. The method according to claim 17, characterized in that, The method further includes: Align the multi-source sensing information in terms of both time and space dimensions; The time dimension alignment includes: determining a reference timestamp; obtaining the original data of the preceding and following frames adjacent to the target timestamp; and calculating the pre-time alignment data of the target timestamp based on the original data of the preceding and following frames using an interpolation method. The spatial dimension alignment includes: establishing a vehicle coordinate system; and projecting radar data and camera data into the vehicle coordinate system.

19. The method according to any one of claims 1 to 3, characterized in that, The step of identifying the current operating condition of the vehicle based on multi-source perception information includes: The multi-source sensing information is input into a pre-trained working condition recognition model, and the temperature, adhesion coefficient and vehicle state are fuzzified and mapped to fuzzy sets; the fuzzy sets include the fuzzy types corresponding to the multi-source sensing information respectively; The fuzzy set is matched with the antecedent conditions of candidate rules in the rule base, the membership degree between the fuzzy type in the fuzzy set and the antecedent conditions is calculated, and the activation strength of the candidate rule is calculated based on the membership degree; the activation strength is used to indicate the degree of matching between the fuzzy set and the antecedent conditions. The target rule is determined based on the activation intensity of the candidate rule, and the working condition corresponding to the target rule is output as the current working condition.

20. The method according to any one of claims 1 to 3, characterized in that, The method further includes: Activate the corresponding management sub-strategy based on the current operating condition; In the case of the current working condition being a complex condition of low temperature and low adhesion, the management sub-strategy includes: limiting the rise rate of the motor peak torque request to within a preset safety envelope; reducing the kinetic energy recovery intensity; sending a command to the chassis domain controller to increase the damping and controlling the ESP to actively establish the brake master cylinder pre-pressure.

21. The method according to any one of claims 1 to 3, characterized in that, The method further includes: Based on the vehicle's current real-time current, battery state of charge (SOC), and internal resistance, the trajectory of terminal voltage changes in the future period is deduced. If the predicted terminal voltage in the terminal voltage change trajectory reaches the undervoltage lower limit, a power cap warning is sent to the vehicle control unit (VCU); the VCU is used to perform forward filtering on the accelerator pedal opening in response to the power cap warning.

22. The method according to any one of claims 1 to 3, characterized in that, The method further includes: The system invokes a neural network model or a magic formula tire model to predict the vehicle's expected slip ratio based on the current expected torque. When the expected slip ratio reaches or exceeds the optimal slip ratio, a first-order inertial low-pass filter is superimposed on the torque request front end; wherein the time constant of the first-order inertial low-pass filter is dynamically adjusted according to the expected slip ratio.

23. An energy management device, characterized in that, The device includes: The identification module is used to identify the current operating condition of the vehicle based on the vehicle's multi-source perception information; The adjustment module is used to adjust the objective function of energy management and the constraints of vehicle control according to the current operating conditions. The objective function is used to balance the economy of energy management, battery temperature and driving safety. The generation module is used to generate control instructions based on the objective function and the constraints, and the control instructions are used to implement energy management that conforms to the objective function and the constraints.

24. A computer device, characterized in that, The computer device includes a processor and a memory, the memory storing at least one program, which is loaded and executed by the processor to implement the energy management method as described in any one of claims 1 to 22.

25. A computer-readable storage medium, characterized in that, The readable storage medium stores at least one program, which is loaded and executed by a processor to implement the energy management method as described in any one of claims 1 to 22.

26. A computer program product, characterized in that, The computer program product includes computer instructions stored in a computer-readable storage medium, a processor of a computer device reading the computer instructions from the computer-readable storage medium, and the processor executing the computer instructions to cause the computer device to perform the energy management method as described in any one of claims 1 to 22.

Images