Dual-motor energy consumption optimization control method and system based on front road predictive driving

Abstract

The invention relates to the field of automobile motor operation modes and energy consumption, in particular to a double-motor energy consumption optimization control method and system based on front road predictive driving, and the method comprises the steps: obtaining the front road information, the vehicle state and the historical operation data of a driver in real time, and predicting the vehicle demand torque at the next moment through the pre-trained neural network model. And comparing the required torque with the maximum torque of the single motor at the current rotating speed, and intelligently generating a plurality of candidate working modes including cooperation of the single motor and the double motors. Based on the motor efficiency MAP, accurately calculating the total input power and the optimal torque distribution result of the system when the torque requirement is met; and for a dual-motor mode, torque distribution between the two motors is optimized in real time by taking the minimum total input power of the system as a target. The total input power of all the candidate modes is compared, the torque distribution scheme with the lowest energy consumption is selected to generate a control instruction, the control instruction is sent to a motor controller to be executed, and therefore prospective energy consumption optimal control is achieved.

Description

Technical Field

[0001] This application relates to the field of automotive motor operation modes and energy consumption, specifically to a dual-motor energy consumption optimization control method and system based on forward-looking driving. Background Technology

[0002] With the acceleration of global energy transition and the electrification of the automotive industry, new energy vehicles, represented by pure electric vehicles, have become an important direction for industrial development. Short driving range and low energy efficiency are among the core technological bottlenecks currently hindering their large-scale adoption and improvement of user experience. For vehicles using a dual-motor drive architecture, they theoretically have the potential to optimize energy efficiency through intelligent torque distribution, but existing technologies still have significant shortcomings in practical applications.

[0003] Currently, most mainstream dual-motor control strategies rely on responsive torque distribution based on the vehicle's instantaneous state, lacking foresight in their decision-making. This results in the system's inability to anticipate and adapt to changes in road conditions ahead, often leading to torque distribution lag and frequent deviations of the motor's operating point from its efficient range when facing complex conditions such as inclines and curves. Furthermore, existing solutions typically rely on fixed rules or simplified efficiency models for mode switching and torque distribution, failing to make globally optimal energy efficiency decisions between single-motor drive and dual-motor cooperative drive. Consequently, the potential of the dual-motor system is not fully utilized, resulting in relatively high overall energy consumption during vehicle operation, unnecessary energy waste, and directly limiting further improvements in vehicle range. Summary of the Invention

[0004] To address the aforementioned problems, this invention provides a dual-motor energy consumption optimization control method and system based on forward-looking driving.

[0005] In a first aspect, the present invention provides a dual-motor energy consumption control method based on predictive driving conditions of the road ahead, comprising the following steps: S1. Obtain road information ahead of the vehicle, and simultaneously obtain real-time vehicle status data and historical driving operation data; input the road information ahead, real-time vehicle status data and historical driving operation data into a pre-trained neural network model to predict the vehicle's required torque at the next moment. S2. Compare the required torque of the vehicle with the maximum output torque of a single motor at the current speed, and generate one or more candidate motor operating modes based on the comparison results; the candidate motor operating modes include single motor candidate mode and dual motor cooperative candidate mode; S3. For each candidate motor operating mode, based on the motor efficiency MAP, calculate the total input power of the vehicle drive system and the corresponding torque distribution result when the required torque of the vehicle is met in this mode; wherein, for the dual-motor cooperative candidate mode, the calculation process takes the minimum total input power as the optimization objective and performs real-time optimization and allocation of the output torque of the two motors. S4. Compare the total input power corresponding to each candidate motor operating mode, select the torque distribution result with the lowest total input power as the final control scheme, and generate motor torque distribution control instructions based on the control scheme, and send them to the vehicle motor controller for execution.

[0006] By integrating road information, real-time vehicle status, and historical driving data, and using a neural network model to accurately predict vehicle torque demand, this system achieves predictive performance control, preventing energy waste caused by delayed torque demand assessments. This solves the core problem of existing dual-motor vehicles' inability to accurately select motor load-bearing modes based on road conditions and load. Based on a comparison of demand torque and the maximum torque of a single motor, it dynamically generates candidate operating modes such as single-motor and dual-motor collaboration. Combining this with a motor efficiency MAP, it calculates the total input power and optimizes torque distribution with the goal of minimizing total energy consumption, ensuring the motor always operates within its high-efficiency range, reducing vehicle energy consumption, and extending driving range.

[0007] As a further limitation of the technical solution of the present invention, step S1 includes: S11. Obtain road feature information ahead of the vehicle through the positioning system and high-precision map; obtain real-time vehicle status data through the vehicle bus; obtain historical driving operation data sequences; S12. The acquired road feature information, vehicle status data and historical driving operation data sequence are input into a pre-trained driving behavior prediction neural network model to predict the rate of change of the accelerator pedal opening of the driver in the next control cycle. S13. Based on the predicted rate of change of accelerator pedal opening and the current pedal opening, calculate the expected pedal opening; input the expected pedal opening, current vehicle speed, vehicle mass, and the immediate road gradient extracted from the road feature information into the vehicle longitudinal dynamics model; calculate the vehicle's required torque; wherein, the expected pedal opening is calculated. The formula is:

[0008] in, The current pedal opening. For the predicted rate of change of the pedal, To control the cycle.

[0009] By clarifying the specific channels for obtaining road feature information, vehicle status data, and historical driving data, and accurately outputting the rate of change of accelerator pedal opening through a driving behavior prediction model, a reliable input is provided for subsequent torque demand calculation, thereby improving the accuracy of torque prediction.

[0010] A clear formula for calculating the expected pedal opening is provided, and the required torque is derived by combining the vehicle's longitudinal dynamics model. The calculation process is quantified to avoid errors caused by empirical judgment, making the torque demand calculation more scientific and repeatable. A control cycle parameter is introduced to adapt to the timing requirements of real-time vehicle control, ensuring that the predicted results are synchronized with the actual driving scenario and improving the real-time response capability of the control scheme.

[0011] As a further limitation of the technical solution of the present invention, the required torque of the vehicle is calculated based on the vehicle's longitudinal dynamics model. The formula:

[0012] in, The radius of the wheel's rolling motion. This refers to the gear ratio of the transmission. Main reducer transmission ratio, The total mechanical efficiency of the transmission system. air density, The vehicle's frontal area. The air drag coefficient, The rolling resistance coefficient, The desired acceleration; the desired acceleration Based on the expected pedal opening and current vehicle speed The desired acceleration is obtained by querying a preset pedal opening-vehicle speed-desired acceleration mapping table. This deeply binds the driver's operating intentions with the vehicle's dynamic characteristics, making the calculated required torque more closely match actual driving needs and avoiding increased energy consumption due to excessive or insufficient torque.

[0013] As a further limitation of the technical solution of the present invention, step S2 includes: S21, Determine the required torque for the vehicle With the maximum output torque of a single motor at the current speed A comparison is made; wherein, the maximum output torque of a single motor at the current speed is... It is a real-time value obtained by querying the current actual speed based on the motor's external characteristic curve or peak torque MAP. S22, if ≤ Then the initialization candidate motor operating mode set must include at least: the first motor independently provides The first single-drive mode, the second motor provides independent power. The second single-drive mode, with the first and second motors working together to provide The dual-motor collaborative mode; if > If so, the initialization candidate motor operating mode set will only include the dual-motor cooperative mode; S23. When the candidate motor operating mode set includes a single-drive mode, calculate the individual load capacity of the first motor and the second motor. The load rate at that time; if the load rate of any motor is lower than the preset inefficient range threshold, then the load rate of the first motor and the second motor is calculated. The system's total input power at that time is used to retain the single-motor operating mode with the lowest total system input power in the candidate set as an effective single-drive candidate mode; the inefficiency interval threshold is determined based on the motor efficiency MAP chart and is the lower limit of the load rate corresponding to the motor efficiency being lower than the first predetermined efficiency value.

[0014] The real-time maximum output torque of a single motor is obtained based on the motor's external characteristic curve or peak torque map, making the comparison between the required torque and the maximum torque of a single motor more closely reflect the motor's real-time operating state and avoiding misjudgments caused by static threshold judgments. When the torque demand is low, multiple options such as single-drive and dual-drive are retained; when the torque demand is high, only the dual-drive mode is retained, balancing energy consumption optimization and power performance assurance, and avoiding invalid calculations. An inefficient range threshold is introduced to filter effective single-drive modes, eliminating single-motor solutions that operate inefficiently under low loads, reducing invalid candidate modes, and improving the efficiency of subsequent energy consumption calculations and solution selection.

[0015] As a further limitation of the technical solution of the present invention, step S3 includes: S31. Traverse the set of candidate motor operating modes; for each candidate mode in the set, based on the current motor speed and the vehicle's required torque, query or interpolate the efficiency MAP of the first motor and the second motor to obtain the motor efficiency value under the corresponding operating condition. S32. For the first motor single-drive mode and the second motor single-drive mode, the total system input power... The calculation method is as follows: ,

[0016] in, This refers to the mechanical power output of a single motor. For this motor at a torque of Rotation speed is The efficiency value at the operating point; S33. For the dual-motor cooperative mode, perform the following optimization allocation and calculation process: Establish based on the total system input power This is an optimization problem with the objective of minimizing the torque of the first motor as the decision variable. Second motor torque ; To meet the torque requirements of the vehicle Under the constraints of torque distribution for each motor, the optimization problem is solved to obtain the optimal torque distribution combination. , ); According to the optimal torque distribution combination ( , ) and current speed The corresponding efficiencies were obtained from the efficiency MAPs of the two motors respectively. and And calculate their respective output power. and Then the total input power of the system for:

[0017] S34. Calculate the operating mode of each candidate motor and the total system input power. The torque distribution result is associated and stored; for single-drive mode, the torque distribution result is [ , 0] or [0, For the dual-motor cooperative mode, the torque distribution result is [ , ].

[0018] The specific calculation methods for the total system input power in single-drive and dual-drive modes are clarified. Efficiency values ​​are obtained by querying or interpolating motor efficiency MAP charts to ensure the accuracy of energy consumption calculations and provide reliable data support for selecting the optimal solution. For dual-drive mode, an optimization model that minimizes total input power is established, clarifying decision variables, constraints, and calculation logic. Through torque distribution optimization, the energy consumption during dual-motor collaborative operation is optimized, solving the energy waste problem caused by unreasonable torque distribution during dual-motor collaboration.

[0019] As a further limitation of the technical solution of the present invention, the optimization problem in S33 is modeled as follows: Objective function:

[0020] Constraints:

[0021]

[0022]

[0023] in, and These are functions of torque and speed, determined based on the efficiency MAP diagram.

[0024] As a further limitation of the technical solution of the present invention, step S4 includes: S41. Compare the total system input power calculated and stored for each candidate motor operating mode. The value is used to determine the total input power that minimizes the value. The torque distribution result, which is uniquely associated with the torque distribution, is used to determine the final control scheme. S42. Based on the torque distribution results included in the final control scheme, generate a corresponding motor torque distribution control command; the control command includes at least: a target operating mode identifier and the target torque of the first motor. and the target torque of the second motor ; S43. Perform a safety check on the generated motor torque distribution control command based on the constraints. The check includes: whether the target torque in the command is within the feasible torque range of each motor at its current speed, and whether the sum of the target torques of the two motors is equal to the current vehicle torque requirement. After the verification is successful, the control command is sent to the vehicle motor controller via the vehicle communication network. S44. Within a predetermined time after the control command is issued, monitor the feedback signal from the motor controller to confirm that the actual output torque of the first motor and the second motor matches the target torque. , Whether the following error is within the allowable tolerance range.

[0025] The optimal solution is determined by comparing the total input power of candidate modes, directly targeting the core objective of minimizing energy consumption and ensuring the effectiveness of the control results. The core content of the control commands is clearly defined to ensure the accuracy of command transmission, enabling the motor controller to precisely execute the optimized solution. A safety verification step for the control commands is added to check the feasibility of the target torque and the matching of the sum of torques, avoiding safety issues such as motor overload and insufficient power due to command errors, thus improving the reliability of the control scheme. A feedback signal monitoring mechanism is introduced to verify the following error between the actual motor output and the target torque in real time, ensuring the execution effect of the control commands, promptly detecting and correcting execution deviations, and guaranteeing energy consumption optimization and driving stability.

[0026] As a further limitation of the technical solution of the present invention, in S41, if there are two or more candidate motor operating modes corresponding to the total system input power When the values ​​are equal and both are the minimum, the final control scheme is selected according to the preset priority rules. The priority rules include: prioritizing the single motor working mode or prioritizing the mode in which the currently working motor continues to work.

[0027] As a further limitation of the technical solution of the present invention, in S44, if the following error of any motor is monitored to exceed the allowable tolerance range for a continuous set time, the fault diagnosis process is triggered, and the system automatically switches to the predefined redundant control mode or limp home mode.

[0028] Secondly, the technical solution of the present invention also provides a dual-motor energy consumption optimization control system based on forward-looking driving, comprising: The road information acquisition module is used to acquire road feature information ahead of the vehicle; The vehicle status acquisition module is used to acquire real-time vehicle status data through the vehicle bus. The driving operation recording module is used to acquire historical driving operation data sequences; The driving intention prediction module is connected to the road information acquisition module, vehicle status acquisition module and driving operation recording module. It is used to input the road feature information, vehicle status data and historical operation data sequence into the pre-trained driving behavior prediction neural network model to predict the rate of change of the driver's accelerator pedal opening in the next control cycle. The demand torque calculation module is connected to the driving intention prediction module. It is used to calculate the expected pedal opening based on the predicted rate of change of accelerator pedal opening and the current pedal opening. The expected pedal opening, current vehicle speed, vehicle mass and the road slope ahead extracted from the road feature information are input into the vehicle longitudinal dynamics model to calculate the vehicle demand torque. The working mode screening module is connected to the required torque calculation module. It is used to compare the required torque of the vehicle with the maximum output torque of a single motor at the current speed, and generate one or more candidate motor working modes based on the comparison results. The energy consumption calculation and optimization module is connected to the working mode screening module. It is used to calculate the total input power of the vehicle drive system and the corresponding torque distribution result when the required torque of the vehicle is met, based on the efficiency MAP of the motor for each candidate motor working mode. The optimal decision execution module, connected to the energy consumption calculation and optimization module, is used to compare the total input power corresponding to each candidate motor working mode, select the torque distribution result with the lowest total input power as the final control scheme, and generate a motor torque distribution control command to send to the vehicle motor controller.

[0029] As a further limitation of the technical solution of the present invention, it also includes: The efficiency MAP storage unit is used to store the efficiency MAP data of the first motor and the second motor. The driver model storage unit is used to store the mapping table of pedal opening, vehicle speed, and desired acceleration. A safety verification unit, connected to the optimal decision execution module, is used to perform safety verification on the generated motor torque distribution control command. The status monitoring unit is used to monitor the error in the actual output torque of the motor and the control command.

[0030] Thirdly, the present invention also provides a vehicle, including a dual-motor energy consumption optimization control system based on forward-looking driving as described in the first aspect.

[0031] As can be seen from the above technical solutions, this application has the following advantages: by integrating information about the road ahead, real-time vehicle status, and historical driving behavior, it achieves intelligent prediction of future vehicle torque demand, and based on this prediction, dynamically selects the torque distribution scheme with optimal energy consumption among various motor operating modes. This solution fundamentally solves the energy waste problem caused by improper load distribution in dual-motor vehicles under complex road conditions, reduces overall vehicle energy consumption, extends the driving range of electric vehicles, and improves the operating efficiency of the power system. Attached Figure Description

[0032] To more clearly illustrate the technical solution of this application, the accompanying drawings used in the description 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.

[0033] Figure 1 This is a flowchart illustrating the method provided in an embodiment of the present invention.

[0034] Figure 2 This is a flowchart of the dual-motor energy consumption control logic in an embodiment of the present invention.

[0035] Figure 3 A block diagram of a system provided in an embodiment of the present invention. Detailed Implementation

[0036] To make the purpose, features, and advantages of this application more apparent and understandable, specific embodiments and accompanying drawings will be used to clearly and completely describe the technical solution protected by this application. Obviously, the embodiments described below are only some embodiments of this application, and not all embodiments. Based on the embodiments in this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0037] Unless otherwise defined, all technical and scientific terms used in this application have the same meaning as commonly understood by one of ordinary skill in the art to which this invention pertains. The terminology used in this application and in the specification of this invention is for the purpose of describing particular embodiments only and is not intended to be limiting of the invention.

[0038] like Figure 1 As shown, this embodiment of the invention provides a dual-motor energy consumption control method based on predictive driving conditions of the road ahead, including the following steps: S1. Acquire road information ahead of the vehicle, and simultaneously acquire real-time vehicle status data and historical driving operation data; input the road information ahead, real-time vehicle status data, and historical driving operation data into a pre-trained neural network model to predict the vehicle's required torque at the next moment; specifically including: S11. Obtain road feature information ahead of the vehicle through the positioning system and high-precision map; obtain real-time vehicle status data through the vehicle bus; obtain historical driving operation data sequences; S12. The acquired road feature information, vehicle status data and historical driving operation data sequence are input into a pre-trained driving behavior prediction neural network model to predict the rate of change of the accelerator pedal opening of the driver in the next control cycle. S13. Based on the predicted rate of change of accelerator pedal opening and the current pedal opening, calculate the expected pedal opening; input the expected pedal opening, current vehicle speed, vehicle mass, and the immediate road gradient extracted from the road feature information into the vehicle longitudinal dynamics model; calculate the vehicle's required torque; wherein, the expected pedal opening is calculated. The formula is:

[0039] in, The current pedal opening. For the predicted rate of change of the pedal, To control the cycle.

[0040] It should be noted that the pre-trained neural network model is a temporal predictive neural network model, which employs a recurrent neural network containing a long short-term memory network structure or a gated recurrent unit structure. The model's input layer is designed to receive data organized by time steps: for historical operation sequences, it is input in chronological order; for forward road sequences, it is input in future time order. This structure enables the model to learn and memorize the dynamic correlation between driving patterns and road conditions over long time spans.

[0041] The model is trained using supervised learning. A real driving dataset is collected, including vehicle state time series, corresponding high-precision map information time series, and driver pedal operation time series. Data from consecutive time segments is used to construct samples, with the model input being... The fusion of temporal features at specific moments, labeled as The actual rate of change of pedal opening at any given time. Backpropagation is used to optimize the model parameters using a time-based algorithm until the prediction accuracy meets the requirements. The specific training process is as follows: Collect a large amount of natural driving data in real vehicles or high-fidelity simulators, covering different roads (highways, cities, mountain roads), different drivers, and different traffic conditions.

[0042] For each sampling time point Construct a fused feature vector Its dimensions are fixed and include: Road feature sub-vector: Predicted road slope sequence within a future T-second time window Curvature sequence, speed limit sequence.

[0043] Vehicle state subvector: current time step speed Longitudinal acceleration Vehicle quality .

[0044] Historical operation subvector: the time sequence of accelerator pedal opening over the past N seconds. Brake pedal signal timing sequence.

[0045] The corresponding label is the actual rate of change of throttle pedal opening in the next control cycle. This value is calculated differentially from the actual collected pedal opening data: .

[0046] The driving behavior prediction neural network model preferably uses a long short-term memory network as its core structure to effectively handle time-dependent operation sequences and road sequences.

[0047] The model input layer receives the above fused feature vectors. .

[0048] The model's output layer consists of one or more neurons that directly output pairs of neurons. The predicted value.

[0049] The preprocessed massive dataset is divided into training, validation, and test sets in chronological order. Supervised learning is employed, using mean squared error or smoothed L1 loss as the loss function to measure the rate of change in the model's predictions. With real labels The differences between them are analyzed. The model parameters are iteratively optimized using the backpropagation algorithm and the Adam optimizer, while performance is monitored on the validation set to prevent overfitting. Training is considered complete when the model performs well on the test set. The predictions achieved the preset accuracy requirements.

[0050] The hidden layers (especially the recurrent layers) inside the model first extract deep features from each part of the input. For example, from the time sequence of accelerator pedal opening. In this model, the driver's operating style and reaction habits can be extracted from the predicted road slope sequence. In this model, the vehicle can understand the terrain challenges it will face (such as needing to refuel when going uphill and potentially reducing throttle when going downhill). This is based on the current vehicle state. , In this model, the vehicle's real-time dynamics can be perceived.

[0051] The model fuses and interacts with the extracted features in a high-dimensional space. For example, it associates the feature of a steep uphill ahead with features such as high current speed and smooth driver history. During training, the model inherently establishes association rules: when an uphill ahead is detected and the current speed is decreasing, and the driver is accustomed to smooth driving, the driver is very likely to slowly and smoothly increase the pedal opening in the following steps.

[0052] Ultimately, the model's output layer calculates a scalar value based on this fused understanding of the driver-vehicle-road integrated system: the predicted rate of change of accelerator pedal opening. This value quantifies the direction and magnitude of the driver's most likely adjustment in the next moment, as determined by the model. Positive values ​​represent accelerating, negative values ​​represent decelerating, and the absolute value represents the speed of the adjustment.

[0053] The required torque of the vehicle is calculated based on the vehicle's longitudinal dynamics model. The formula:

[0054] in, The radius of the wheel's rolling motion. This refers to the gear ratio of the transmission. Main reducer transmission ratio, The total mechanical efficiency of the transmission system. air density, The vehicle's frontal area. The air drag coefficient, The rolling resistance coefficient, The desired acceleration; the desired acceleration Based on the expected pedal opening and current vehicle speed The preset pedal opening-vehicle speed-desired acceleration mapping relationship table is obtained by querying it.

[0055] The pedal opening-vehicle speed-desired acceleration mapping table, also known as the driver demand model or driving force characteristic spectrum, is a systematic calibration process that integrates subjective driving performance evaluation, objective vehicle testing, and engineering optimization. Its goal is to map discrete driver operations (pedal opening) and vehicle states (vehicle speed) into continuous and smooth acceleration requests that meet the driver's psychological expectations.

[0056] The core of establishing this mapping relationship is simulating the expected acceleration capability of a vehicle from an ideal driver at a given pedal depth and vehicle speed. This requires simultaneously satisfying: The deeper you press the pedal, the greater the expected acceleration (at the same vehicle speed).

[0057] The higher the vehicle speed, the smaller the acceleration that can be achieved with the same pedal depth (limited by motor power).

[0058] The mapping relationship is smooth, avoiding abrupt changes in acceleration that could cause a sense of jerkiness.

[0059] Different mapping tables can be created for different driving modes (such as Eco, Comfort, and Sport).

[0060] The specific steps for setting it up are as follows: Step 1: Based on the peak torque characteristics of the drive motor and the vehicle parameters, calculate the maximum theoretical acceleration corresponding to each vehicle speed under full throttle conditions. This serves as the physical upper limit of the mapping relationship, within the pedal opening range [0%, 100%] and the vehicle speed range [0, 100%]. Within [the specified range], multiple calibration operating points are selected at predetermined intervals. ; Based on the peak torque-speed external characteristic curve of the drive motor and the vehicle parameters, the maximum physical acceleration achievable at each vehicle speed point under full throttle (100% pedal opening) is calculated. This is the upper limit of the mapping table. The minimum expected acceleration typically corresponds to the coasting or regenerative braking acceleration when the pedal opening is 0%.

[0061] exist On a two-dimensional plane, a grid is formed with one point for every 10% of pedal opening and one point for every 10 km / h of vehicle speed, creating a series of operating points to be calibrated. .

[0062] Step Two: In the real-vehicle test, the evaluation driver will test the vehicle at a constant speed. Next, perform the stepping motion to the target opening degree. The calibration engineer iteratively adjusts and ultimately determines the operating point based on the driver's subjective feedback regarding excessive, weak, or appropriate acceleration response. The optimal expected acceleration value ; On a flat, dry, and good road surface, maintain a constant speed. Cruise control. Evaluate the driver's ability to quickly and steadily depress the pedals to the target opening degree. , and maintain.

[0063] The driver focuses on experiencing the initial response speed and continuous acceleration of the vehicle as it accelerates from its current constant speed, judging whether it is natural or expected.

[0064] The calibration engineer adjusts the corresponding settings in the background. Initial value of desired acceleration at the operating point The driver experiences the acceleration again, providing feedback on whether it's too strong, too weak, or just right. After multiple iterations, the acceleration at that specific operating point is evaluated as optimal. At this point, the code uses... It was then decided.

[0065] Step 3: Apply the calibration results to all the discrete points obtained. A bilinear interpolation algorithm is used to generate a continuous and smooth desired acceleration response surface covering the entire range of pedal opening and vehicle speed. ; Step two yielded a series of discrete operating points. The value. Using bilinear interpolation or surface fitting algorithms (such as two-dimensional spline interpolation), in all... Generate a continuous and smooth desired acceleration response surface on the plane. Ensure the generated surface is monotonically increasing at all operating points (as the curve changes). (Increases as needed), but does not exceed the physical upper limit determined in step one. .

[0066] Step 4: By changing the target set by the subjective evaluation, repeat steps 2 and 3 to generate multiple mapping tables corresponding to different driving modes; Economic model: in the same Below, a relatively lower value is determined. Encourage gentle driving to reduce energy consumption.

[0067] Sport mode: calibrated for a more aggressive and faster response In some operating conditions, values ​​closer to the physical upper limit are even used.

[0068] Finally, a set of mapping tables corresponding to different driving modes is generated.

[0069] Step 5: Integration verification and vehicle testing; The calibrated mapping table is written into the vehicle controller, and a comprehensive road test is performed.

[0070] The test verifies whether the vehicle's acceleration response is consistent, linear, and as expected under different road conditions and driver operations, without any abruptness.

[0071] Make minor adjustments based on test feedback.

[0072] The generated mapping table is stored in the vehicle controller as a two-dimensional data table for real-time querying.

[0073] S2. Compare the required torque of the vehicle with the maximum output torque of a single motor at the current speed, and generate one or more candidate motor operating modes based on the comparison results; the candidate motor operating modes include single-motor candidate modes and dual-motor collaborative candidate modes; this step specifically includes: S21, Determine the required torque for the vehicle With the maximum output torque of a single motor at the current speed A comparison is made; wherein, the maximum output torque of a single motor at the current speed is... It is a real-time value obtained by querying the current actual speed based on the motor's external characteristic curve or peak torque MAP. S22, if ≤ Then the initialization candidate motor operating mode set must include at least: the first motor independently provides The first single-drive mode, the second motor provides independent power. The second single-drive mode, with the first and second motors working together to provide The dual-motor collaborative mode; if > If so, the initialization candidate motor operating mode set will only include the dual-motor cooperative mode; S23. When the candidate motor operating mode set includes a single-drive mode, calculate the individual load capacity of the first motor and the second motor. The load rate at that time; if the load rate of any motor is lower than the preset inefficient range threshold, then the load rate of the first motor and the second motor is calculated. The system's total input power at that time is used to retain the single-motor operating mode with the lowest total system input power in the candidate set as an effective single-drive candidate mode; the inefficiency interval threshold is determined based on the motor efficiency MAP chart and is the lower limit of the load rate corresponding to the motor efficiency being lower than the first predetermined efficiency value.

[0074] S3. For each candidate motor operating mode, based on the motor's efficiency MAP, calculate the total input power of the vehicle drive system and the corresponding torque distribution result when the required torque of the vehicle is met in that mode; wherein, for the dual-motor cooperative candidate mode, the calculation process takes the minimum total input power as the optimization objective and performs real-time optimization and allocation of the output torque of the two motors; This step specifically includes: S31. Traverse the set of candidate motor operating modes; for each candidate mode in the set, based on the current motor speed and the vehicle's required torque, query or interpolate the efficiency MAP of the first motor and the second motor to obtain the motor efficiency value under the corresponding operating condition. S32. For the first motor single-drive mode and the second motor single-drive mode, the total system input power... The calculation method is as follows: ,

[0075] in, This refers to the mechanical power output of a single motor. For this motor at a torque of Rotation speed is The efficiency value at the operating point; S33. For the dual-motor cooperative mode, perform the following optimization allocation and calculation process: Establish based on the total system input power This is an optimization problem with the objective of minimizing the torque of the first motor as the decision variable. Second motor torque ; To meet the torque requirements of the vehicle Under the constraints of torque distribution for each motor, the optimization problem is solved to obtain the optimal torque distribution combination. , ); According to the optimal torque distribution combination ( , ) and current speed The corresponding efficiencies were obtained from the efficiency MAPs of the two motors respectively. and And calculate their respective output power. and Then the total input power of the system for:

[0076] The optimization problem is modeled as follows: Objective function:

[0077] Constraints:

[0078]

[0079]

[0080] in, and These are functions of torque and speed, determined based on the efficiency MAP diagram.

[0081] The motor efficiency value Motor torque and rotational speed The function, whose relationship is stored in the controller as an efficiency map (i.e., a two-dimensional data table), allows the determination of any desired function by querying the efficiency map and interpolating the results. The continuous efficiency value corresponding to the operating point.

[0082] S34. Calculate the operating mode of each candidate motor and the total system input power. The torque distribution result is associated and stored; for single-drive mode, the torque distribution result is [ , 0] or [0, For the dual-motor cooperative mode, the torque distribution result is [ , ].

[0083] S4. Compare the total input power corresponding to each candidate motor operating mode, select the torque distribution result with the lowest total input power as the final control scheme, and generate motor torque distribution control commands based on this control scheme, sending them to the vehicle motor controller for execution. Specifically, this includes: S41. Compare the total system input power calculated and stored for each candidate motor operating mode. The value is used to determine the total input power that minimizes the value. The torque distribution result, which is uniquely associated with the motor, is used to determine the final control scheme. In this step, if there are two or more candidate motor operating modes corresponding to the total system input power... When the values ​​are equal and both are the minimum, the final control scheme is selected according to the preset priority rules. The priority rules include: prioritizing the single motor working mode or prioritizing the mode in which the currently working motor continues to work.

[0084] S42. Based on the torque distribution results included in the final control scheme, generate a corresponding motor torque distribution control command; the control command includes at least: a target operating mode identifier and the target torque of the first motor. and the target torque of the second motor ; S43. Perform a safety check on the generated motor torque distribution control command based on the constraints. The check includes: whether the target torque in the command is within the feasible torque range of each motor at its current speed, and whether the sum of the target torques of the two motors is equal to the current vehicle torque requirement. After the verification is successful, the control command is sent to the vehicle motor controller via the vehicle communication network. S44. Within a predetermined time after the control command is issued, monitor the feedback signal from the motor controller to confirm that the actual output torque of the first motor and the second motor matches the target torque. , Check if the following error is within the allowable tolerance range. In this step, if the following error of any motor is monitored to exceed the allowable tolerance range for a continuous set time, the fault diagnosis process is triggered, and the system automatically switches to the predefined redundant control mode or limp-home mode.

[0085] like Figure 2 As shown, the dual-motor energy consumption control logic based on predictive driving conditions of the road ahead in this embodiment of the invention first obtains the vehicle's location information, road conditions ahead, and road resistance coefficient based on GPS positioning and a high-precision three-dimensional map to establish a road model; secondly, it obtains information such as the vehicle's speed, load, pedal opening, motor torque, motor efficiency MAP, remaining battery power, and battery temperature through a CAN device. After obtaining the relevant information, it uses a neural network algorithm to predict the driver's next pedal change rate based on the vehicle's location and the driver's driving habits.

[0086] The next pedal position is obtained based on the predicted pedal change rate (by obtaining the current pedal position and combining it with the predicted pedal change rate, the next pedal position can be obtained). The power required for the vehicle to travel at the current vehicle position is predicted by a neural network algorithm, and the required torque of the vehicle is calculated by combining it with the current speed.

[0087] Determine whether a single motor can carry the full required torque load based on the required torque and the maximum torque of the motor at the current speed. If the maximum torque of the single motor is greater than or equal to the required torque, it can carry the full required torque; otherwise, it cannot.

[0088] If a single motor can handle the full torque, determine if its load rate is within the high-efficiency range. If it cannot, activate the dual-motor mode and optimize torque distribution between the two motors using an optimization method (based on the motor efficiency MAP, with the torque of both motors as decision variables, peak torque / power as constraints, and minimum power as the objective function). Calculate the optimal load rate operating point for each motor and the total power. If the single motor load rate is low, calculate the power of each motor at low load and operate with the motor having the lower power. If the single motor load rate is within the high-efficiency range, calculate the power of the single motor operating alone and the total power when both motors are under the same load. If the single motor load rate is high, activate the dual-motor mode, calculate the optimal load rate operating point for each motor, and the total power. Select the mode with lower energy consumption as the motor operating mode.

[0089] like Figure 3 As shown, this embodiment of the invention also provides a dual-motor energy consumption optimization control system based on forward-looking driving, comprising: The road information acquisition module is used to acquire road feature information ahead of the vehicle; The vehicle status acquisition module is used to acquire real-time vehicle status data through the vehicle bus. The driving operation recording module is used to acquire historical driving operation data sequences; The driving intention prediction module is connected to the road information acquisition module, vehicle status acquisition module and driving operation recording module. It is used to input the road feature information, vehicle status data and historical operation data sequence into the pre-trained driving behavior prediction neural network model to predict the rate of change of the driver's accelerator pedal opening in the next control cycle. The demand torque calculation module is connected to the driving intention prediction module. It is used to calculate the expected pedal opening based on the predicted rate of change of accelerator pedal opening and the current pedal opening. The expected pedal opening, current vehicle speed, vehicle mass and the road slope ahead extracted from the road feature information are input into the vehicle longitudinal dynamics model to calculate the vehicle demand torque. The working mode screening module is connected to the required torque calculation module. It is used to compare the required torque of the vehicle with the maximum output torque of a single motor at the current speed, and generate one or more candidate motor working modes based on the comparison results. The energy consumption calculation and optimization module is connected to the working mode screening module. It is used to calculate the total input power of the vehicle drive system and the corresponding torque distribution result when the required torque of the vehicle is met, based on the efficiency MAP of the motor for each candidate motor working mode. The optimal decision execution module, connected to the energy consumption calculation and optimization module, is used to compare the total input power corresponding to each candidate motor working mode, select the torque distribution result with the lowest total input power as the final control scheme, and generate a motor torque distribution control command to send to the vehicle motor controller.

[0090] In some embodiments, the system further includes: The efficiency MAP storage unit is used to store the efficiency MAP data of the first motor and the second motor. The driver model storage unit is used to store the mapping table of pedal opening, vehicle speed, and desired acceleration. A safety verification unit, connected to the optimal decision execution module, is used to perform safety verification on the generated motor torque distribution control command. The status monitoring unit is used to monitor the error in the actual output torque of the motor and the control command.

[0091] This invention also provides a vehicle including a dual-motor energy consumption optimization control system based on forward-looking driving as described in the above embodiments.

[0092] The above description of the disclosed embodiments enables those skilled in the art to make or use the invention. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the invention. Therefore, the invention is not to be limited to the embodiments shown herein, but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A dual-motor energy consumption control method based on predictive driving conditions of the road ahead, characterized in that, Includes the following steps: S1. Obtain road information ahead of the vehicle, and simultaneously obtain real-time vehicle status data and historical driving operation data; input the road information ahead, real-time vehicle status data and historical driving operation data into a pre-trained neural network model to predict the vehicle's required torque at the next moment. S2. Compare the required torque of the vehicle with the maximum output torque of a single motor at the current speed, and generate one or more candidate motor operating modes based on the comparison results; the candidate motor operating modes include single motor candidate mode and dual motor cooperative candidate mode; S3. For each candidate motor operating mode, based on the motor efficiency MAP, calculate the total input power of the vehicle drive system and the corresponding torque distribution result when the required torque of the vehicle is met in this mode; wherein, for the dual-motor cooperative candidate mode, the calculation process takes the minimum total input power as the optimization objective and performs real-time optimization and allocation of the output torque of the two motors. S4. Compare the total input power corresponding to each candidate motor operating mode, select the torque distribution result with the lowest total input power as the final control scheme, and generate motor torque distribution control instructions based on the control scheme, and send them to the vehicle motor controller for execution.

2. The dual-motor energy consumption control method based on predictive driving conditions of the road ahead, as described in claim 1, is characterized in that, The steps in S1 include: S11. Obtain road feature information ahead of the vehicle through the positioning system and high-precision map; obtain real-time vehicle status data through the vehicle bus; obtain historical driving operation data sequences; S12. Input the acquired road feature information, vehicle status data and historical driving operation data sequence into the pre-trained driving behavior prediction neural network model to predict the rate of change of the driver's accelerator pedal opening in the next control cycle. S13. Based on the predicted rate of change of accelerator pedal opening and the current pedal opening, calculate the expected pedal opening; input the expected pedal opening, current vehicle speed, vehicle mass, and the immediate road gradient extracted from the road feature information into the vehicle longitudinal dynamics model; calculate the vehicle's required torque; wherein, the expected pedal opening is calculated. The formula is: in, The current pedal opening. For the predicted rate of change of the pedal, To control the cycle.

3. The dual-motor energy consumption control method based on predictive driving conditions of the road ahead, as described in claim 2, is characterized in that... The required torque of the vehicle is calculated based on the vehicle's longitudinal dynamics model. The formula: in, The radius of the wheel's rolling motion. This refers to the gear ratio of the transmission. Main reducer transmission ratio, The total mechanical efficiency of the transmission system. air density, The vehicle's frontal area. The air drag coefficient, The rolling resistance coefficient, The desired acceleration; the desired acceleration Based on the expected pedal opening and current vehicle speed The preset pedal opening-vehicle speed-desired acceleration mapping relationship table is obtained by querying it.

4. The dual-motor energy consumption control method based on predictive driving conditions of the road ahead, as described in claim 3, is characterized in that... The steps in S2 include: S21, Determine the required torque for the vehicle With the maximum output torque of a single motor at the current speed A comparison is made; wherein, the maximum output torque of a single motor at the current speed is... It is a real-time value obtained by querying the current actual speed based on the motor's external characteristic curve or peak torque MAP. S22, if ≤ Then the initialization candidate motor operating mode set must include at least: the first motor independently provides The first single-drive mode, the second motor provides independent power. The second single-drive mode, with the first and second motors working together to provide The dual-motor collaborative mode; if > If so, the initialization candidate motor operating mode set will only include the dual-motor cooperative mode; S23. When the candidate motor operating mode set includes a single-drive mode, calculate the individual load capacity of the first motor and the second motor. The load rate at that time; if the load rate of any motor is lower than the preset inefficient range threshold, then the load rate of the first motor and the second motor is calculated. The system's total input power at that time is used to retain the single-motor operating mode with the lowest total system input power in the candidate set as an effective single-drive candidate mode; the inefficiency interval threshold is determined based on the motor efficiency MAP chart and is the lower limit of the load rate corresponding to the motor efficiency being lower than the first predetermined efficiency value.

5. The dual-motor energy consumption control method based on predictive driving conditions of the road ahead, as described in claim 4, is characterized in that... The steps in S3 include: S31. Traverse the set of candidate motor operating modes; for each candidate mode in the set, based on the current motor speed and the vehicle's required torque, query or interpolate the efficiency MAP of the first motor and the second motor to obtain the motor efficiency value under the corresponding operating condition. S32. For the first motor single-drive mode and the second motor single-drive mode, the total system input power... The calculation method is as follows: ， in, This refers to the mechanical power output of a single motor. For this motor at a torque of Rotation speed is The efficiency value at the operating point; S33. For the dual-motor cooperative mode, perform the following optimization allocation and calculation process: Establish based on the total system input power This is an optimization problem with the objective of minimizing the torque of the first motor as the decision variable. Second motor torque ; To meet the torque requirements of the vehicle Under the constraints of torque distribution for each motor, the optimization problem is solved to obtain the optimal torque distribution combination. , ); According to the optimal torque distribution combination ( , ) and current speed The corresponding efficiencies were obtained from the efficiency MAPs of the two motors respectively. and And calculate their respective output power. and Then the total input power of the system for: S34. Calculate the operating mode of each candidate motor and the total system input power. The torque distribution result is associated and stored; for single-drive mode, the torque distribution result is [ , 0] or [0, For the dual-motor cooperative mode, the torque distribution result is [ , ].

6. The dual-motor energy consumption control method based on predictive driving conditions of the road ahead, as described in claim 5, is characterized in that... The optimization problem in S33 is modeled as follows: Objective function: Constraints: in, and These are functions of torque and speed, determined based on the efficiency MAP diagram.

7. The dual-motor energy consumption control method based on predictive driving conditions of the road ahead, as described in claim 6, is characterized in that... The steps in S4 include: S41. Compare the total system input power calculated and stored for each candidate motor operating mode. The value is used to determine the total input power that minimizes the value. The torque distribution result, which is uniquely associated with the torque distribution, is used to determine the final control scheme. S42. Based on the torque distribution results included in the final control scheme, generate a corresponding motor torque distribution control command; the control command includes at least: a target operating mode identifier and the target torque of the first motor. and the target torque of the second motor ; S43. Perform a safety check on the generated motor torque distribution control command based on the constraints. The check includes: whether the target torque in the command is within the feasible torque range of each motor at its current speed, and whether the sum of the target torques of the two motors is equal to the current vehicle torque requirement. After the verification is successful, the control command is sent to the vehicle motor controller via the vehicle communication network. S44. Within a predetermined time after the control command is issued, monitor the feedback signal from the motor controller to confirm that the actual output torque of the first motor and the second motor matches the target torque. , Whether the following error is within the allowable tolerance range.

8. The dual-motor energy consumption control method based on predictive driving conditions of the road ahead, as described in claim 6, is characterized in that... In S41, if there are two or more candidate motor operating modes corresponding to the total system input power When the values ​​are equal and both are the minimum, the final control scheme is selected according to the preset priority rules. The priority rules include: prioritizing the single-motor working mode or prioritizing the mode in which the currently working motor continues to work.

9. The dual-motor energy consumption control method based on predictive driving conditions of the road ahead, as described in claim 6, is characterized in that... In S44, if the tracking error of any motor is monitored to exceed the allowable tolerance range for a continuous set time, the fault diagnosis process is triggered, and the system automatically switches to the predefined redundant control mode or limp home mode.

10. A dual-motor energy consumption optimization control system based on forward-looking driving, characterized in that, include: The road information acquisition module is used to acquire road feature information ahead of the vehicle; The vehicle status acquisition module is used to acquire real-time vehicle status data through the vehicle bus. The driving operation recording module is used to acquire historical driving operation data sequences; The driving intention prediction module is connected to the road information acquisition module, vehicle status acquisition module and driving operation recording module. It is used to input the road feature information, vehicle status data and historical operation data sequence into the pre-trained driving behavior prediction neural network model to predict the rate of change of the driver's accelerator pedal opening in the next control cycle. The demand torque calculation module is connected to the driving intention prediction module. It is used to calculate the expected pedal opening based on the predicted rate of change of accelerator pedal opening and the current pedal opening. The expected pedal opening, current vehicle speed, vehicle mass and the road slope ahead extracted from the road feature information are input into the vehicle longitudinal dynamics model to calculate the vehicle demand torque. The working mode screening module is connected to the required torque calculation module. It is used to compare the required torque of the vehicle with the maximum output torque of a single motor at the current speed, and generate one or more candidate motor working modes based on the comparison results. The energy consumption calculation and optimization module is connected to the working mode screening module. It is used to calculate the total input power of the vehicle drive system and the corresponding torque distribution result when the required torque of the vehicle is met, based on the efficiency MAP of the motor for each candidate motor working mode. The optimal decision execution module, connected to the energy consumption calculation and optimization module, is used to compare the total input power corresponding to each candidate motor working mode, select the torque distribution result with the lowest total input power as the final control scheme, and generate a motor torque distribution control command to send to the vehicle motor controller.

Images