A new energy vehicle electric drive assembly energy consumption optimization method and system

Abstract

The application discloses a new energy automobile electric drive assembly energy consumption optimization method and system, relates to the new energy automobile electric drive assembly energy consumption optimization technical field, and includes: constructing a dynamic space-time topology atlas among household devices; real-time monitoring state data of each node device in the topology atlas; based on the state data and a preset abnormal rule, identifying a source abnormal node in the topology atlas; taking the source abnormal node as input, combining the direction and weight of the edge to execute fault propagation simulation deduction, obtaining a risk device node set and a fault propagation path; based on the set and the propagation path, generating diagnosis early warning information for positioning fault root cause and characterizing propagation link. The application unifies the physical connection and logical dependence of the device through dynamic topology modeling, can quickly lock the root cause and predict the potential affected devices when multiple devices are abnormal at the same time, realizes the forward-looking early warning of implicit chain failure, and improves the operation and maintenance efficiency and system reliability through a visual report.

Description

Technical Field

[0001] This invention relates to the field of electric drive system control technology for new energy vehicles, specifically to a method and system for optimizing the energy consumption of electric drive assemblies for new energy vehicles. Background Technology

[0002] With the rapid development of the new energy vehicle industry, the electric drive system, as a core component of the vehicle's powertrain, directly impacts the vehicle's range and energy utilization. Current new energy vehicle electric drive systems mostly employ permanent magnet synchronous motors combined with mature control strategies such as vector control and FOC control, continuously improving efficiency in inverter, power device, and motor design. Simultaneously, vehicle domain controllers are gradually gaining the ability to coordinate control of the drive motor, battery, and thermal management system. Supported by in-vehicle navigation, high-precision maps, and vehicle-to-everything (V2X) technology, vehicles can already acquire some road and traffic information ahead, enabling functions such as cruise control and energy recovery control, laying the hardware and information foundation for higher-level vehicle energy consumption optimization.

[0003] However, existing methods for optimizing the energy consumption of electric drive assemblies in new energy vehicles still have several limitations. On the one hand, existing solutions mostly adopt feedback control or empirical rule-based energy management based on instantaneous operating conditions. Although they can adjust torque output based on information such as current vehicle speed and pedal opening, they are difficult to use predictive information such as road slope, curvature, and traffic light phase to predict the power demand for future driving cycles in a timely and systematic manner. They cannot optimize vehicle speed trajectory and torque demand at the "travel level," resulting in limited room for optimization of the overall vehicle drive energy consumption. On the other hand, the control of permanent magnet synchronous motors usually relies on fixed efficiency maps and MTPA / MTPV curves calibrated offline during the production or testing phase. They fail to fully consider the real-time drift of electromagnetic parameters such as motor inductance and flux linkage with temperature, saturation level, and aging state. This causes the "theoretical high efficiency point" given by the controller according to the static baseline efficiency mapping curve to often deviate from the motor's true maximum efficiency condition in actual operation, resulting in efficiency loss over long-term operation. Furthermore, for electric drive assemblies employing dual-motor or multi-motor structures, existing torque distribution strategies are mostly fixed ratios or simple rules, lacking a global comprehensive efficiency-optimized allocation mechanism based on real-time corrected efficiency mapping curves. This makes it difficult to simultaneously consider the efficiency differences between the front and rear motors and real-time operating condition changes. Therefore, existing technologies cannot yet achieve integrated energy consumption optimization control of the electric drive assembly under complex road and traffic conditions, comprehensively utilizing route prediction information, online parameter identification results of permanent magnet synchronous motors, and multi-motor collaborative control. Summary of the Invention

[0004] In view of the above-mentioned problems, the present invention is proposed.

[0005] Therefore, the technical problems solved by this invention are: existing energy consumption optimization methods for electric drive assemblies of new energy vehicles have the problem of difficulty in timely and effective use of road and traffic information for predictive power planning; permanent magnet synchronous motor control relies on static efficiency mapping curves and is difficult to adapt to real-time drift of motor electromagnetic parameters; torque distribution of multi-motor electric drive assemblies fails to achieve optimal comprehensive efficiency based on real-time efficiency characteristics; and how to achieve integrated energy consumption optimization of electric drive assemblies that combines route prediction, online parameter identification, and multi-motor coordinated torque control under complex driving conditions.

[0006] To solve the above-mentioned technical problems, the present invention provides the following technical solution:

[0007] In a first aspect, embodiments of the present invention provide a method for optimizing the energy consumption of an electric drive assembly for a new energy vehicle. The electric drive assembly includes at least two permanent magnet synchronous motors and their controllers, and the controllers are pre-set with a reference efficiency mapping curve for the permanent magnet synchronous motors. The method is characterized by the following steps:

[0008] Obtain road and traffic information for the vehicle's preset route ahead;

[0009] Based on the road and traffic information, predict the vehicle power demand during future driving cycles and generate a total demand torque command sequence accordingly.

[0010] During the current operation of the vehicle, the permanent magnet synchronous motor is identified online in real time to obtain the real-time electromagnetic parameters of the permanent magnet synchronous motor.

[0011] The reference efficiency mapping curve is dynamically corrected based on the real-time electromagnetic parameters.

[0012] Based on the revised baseline efficiency mapping curve, the total demand torque command sequence is decomposed and synthesized into real-time control commands for each permanent magnet synchronous motor.

[0013] The real-time control command is executed to drive the permanent magnet synchronous motor to run.

[0014] As a preferred embodiment of the energy consumption optimization method for the electric drive assembly of new energy vehicles described in this invention, the step of obtaining road and traffic information for the preset journey ahead of the vehicle includes:

[0015] The vehicle navigation system uses electronic maps to obtain information on road slope and curvature.

[0016] Real-time traffic information is obtained from road infrastructure and traffic management systems through vehicle-to-everything (V2X) communication. This real-time traffic information includes traffic light status, phase timing, traffic flow congestion status, and speed limit information.

[0017] As a preferred embodiment of the energy consumption optimization method for the electric drive assembly of new energy vehicles described in this invention, the step of generating the total demand torque command sequence includes:

[0018] Based on road slope, road curvature and road speed limit information, a longitudinal dynamics model of the vehicle is established in a preset prediction time domain, and the traffic light status and phase sequence are used as vehicle speed constraints.

[0019] With the goal of minimizing the vehicle's driving energy consumption in the predicted time domain, a model predictive control algorithm is used to perform rolling optimization on the vehicle's longitudinal dynamics model to obtain a reference vehicle speed sequence for future driving cycles.

[0020] Based on the reference vehicle speed sequence, road gradient information, and the vehicle longitudinal dynamics model, the total demand torque of the vehicle at each moment required to follow the reference vehicle speed sequence is calculated, and the total demand torque command sequence is formed from the total demand torque of the vehicle at each moment.

[0021] As a preferred embodiment of the energy consumption optimization method for the electric drive assembly of new energy vehicles described in this invention, the step of performing online real-time parameter identification includes:

[0022] During the drive control cycle of the permanent magnet synchronous motor, a preset high-frequency voltage signal is superimposed on the fundamental voltage vector output by the inverter to excite the generation of a high-frequency response current containing motor inductance information.

[0023] The three-phase current of the synchronously sampled motor is separated from the three-phase current by coordinate transformation and bandpass filtering to extract the current response component that is in the same frequency as the high-frequency voltage signal.

[0024] Based on the amplitude and frequency of the high-frequency voltage signal and the separated current response components, the direct-axis inductance and quadrature-axis inductance of the permanent magnet synchronous motor are calculated in real time using the high-frequency impedance model of the permanent magnet synchronous motor, and are used as the real-time electromagnetic parameters.

[0025] As a preferred embodiment of the energy consumption optimization method for the electric drive assembly of new energy vehicles according to the present invention, the dynamic correction of the benchmark efficiency mapping curve includes:

[0026] Obtain the baseline efficiency mapping curve and its associated motor parameterized mathematical model, wherein the motor parameterized mathematical model defines the mapping relationship between the motor electromagnetic parameters and the optimal efficiency operating point of the permanent magnet synchronous motor.

[0027] Substitute the direct-axis inductance and quadrature-axis inductance from the real-time electromagnetic parameters into the parameterized mathematical model of the motor, and update the model parameters in the mapping relationship;

[0028] Based on the updated mapping relationship, the baseline efficiency mapping curve is recalculated or interpolated to generate a corrected efficiency mapping curve that reflects the actual characteristics of the current motor.

[0029] As a preferred embodiment of the energy consumption optimization method for the electric drive assembly of new energy vehicles described in this invention, the step of decomposing and synthesizing the total demand torque command sequence into real-time control commands for each permanent magnet synchronous motor includes:

[0030] Based on the total torque demand command at the current moment, the torque allocation command for each permanent magnet synchronous motor is determined according to the principle of optimizing the real-time comprehensive efficiency of the electric drive assembly based on the modified efficiency mapping curve.

[0031] Based on the torque distribution command and current speed of each permanent magnet synchronous motor, the corresponding corrected efficiency mapping curve is queried to obtain the optimal efficiency current command of each permanent magnet synchronous motor under the corresponding operating conditions.

[0032] The optimal efficiency current command is used as the real-time control command.

[0033] As a preferred embodiment of the energy consumption optimization method for the electric drive assembly of new energy vehicles according to the present invention, the execution of the real-time control command includes:

[0034] Based on the current command in the real-time control command, a voltage command is generated in the current loop controller;

[0035] Based on the current DC bus voltage, real-time speed and voltage command of each permanent magnet synchronous motor, the pulse width modulation strategy of the inverter is dynamically adjusted.

[0036] Based on the pulse width modulation strategy, a corresponding switching signal for the power switching device is generated, which drives the power switching device of the inverter to operate, so that each permanent magnet synchronous motor outputs a drive torque corresponding to the torque distribution command.

[0037] Secondly, embodiments of the present invention provide an energy consumption optimization system for electric drive assemblies of new energy vehicles, including:

[0038] Information acquisition module: Acquires road and traffic information for the vehicle's preset route ahead;

[0039] Torque planning module: Based on the road and traffic information, predicts the vehicle's power demand during future driving cycles and generates a total demand torque command sequence accordingly;

[0040] Parameter identification module: During the current operation of the vehicle, the permanent magnet synchronous motor is identified online in real time to obtain the real-time electromagnetic parameters of the permanent magnet synchronous motor;

[0041] Curve correction module: Dynamically corrects the reference efficiency mapping curve based on the real-time electromagnetic parameters;

[0042] Command synthesis module: Based on the corrected benchmark efficiency mapping curve, the total demand torque command sequence is decomposed and synthesized into real-time control commands for each permanent magnet synchronous motor.

[0043] Drive control module: Executes the real-time control commands to drive the permanent magnet synchronous motor.

[0044] The beneficial effects of this invention are as follows: By integrating vehicle-road cooperative forward prediction with online real-time parameter identification of the electric drive system, this invention dynamically corrects the motor efficiency mapping curve and uses this as a basis for torque optimization allocation and control execution, thereby achieving closed-loop collaborative optimization of the electric drive assembly's energy consumption. This significantly improves the overall integrity and adaptability of energy consumption optimization, solves the problem of prediction model mismatch caused by time-varying motor parameters, and enables the system to track the true optimal efficiency point in real time under any operating condition. It achieves consistency from global planning to local execution, ensuring that road condition-based energy-saving driving strategies can be accurately executed. It also taps into the global energy efficiency potential of multi-motor systems, further reducing system losses through dynamic optimal efficiency allocation. Ultimately, this method can achieve more thorough and robust energy-saving effects in complex and ever-changing real-world driving environments, effectively extending the vehicle's driving range. Attached Figure Description

[0045] To more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings used in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort, wherein:

[0046] Figure 1 The first embodiment of the present invention provides an overall flowchart of a method for optimizing the energy consumption of an electric drive assembly for new energy vehicles. Detailed Implementation

[0047] To make the above-mentioned objects, features, and advantages of the present invention more apparent and understandable, specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of the present invention, and not all of them. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort should fall within the protection scope of the present invention.

[0048] Example 1, referring to Figure 1In one embodiment of the present invention, a method for optimizing the energy consumption of an electric drive assembly for a new energy vehicle is provided, wherein the electric drive assembly includes at least two permanent magnet synchronous motors and their controllers, and the controllers are pre-set with a reference efficiency mapping curve for the permanent magnet synchronous motors.

[0049] In this embodiment, the preset "baseline efficiency mapping curve" in the controller can be understood as one or more efficiency mapping maps constructed offline based on test data of the permanent magnet synchronous motor under different speed and output torque conditions during the motor factory calibration stage. Preferably, the mapping map is stored in the memory of the motor controller or electric drive domain controller in a two-dimensional lookup table manner, with the horizontal axis representing the motor speed and the vertical axis representing the motor output torque. The table entry corresponding to each grid point may include the motor efficiency value under that condition, and / or the target control parameters that enable the motor to achieve a high-efficiency operating state under that condition, such as the components of the current vector in the direct and quadrature axes, field weakening control-related parameters, etc. The baseline efficiency mapping curve can be obtained offline through bench testing, simulation calculation, or a combination of both, and is fixed in the controller before the vehicle is put into use as the initial basis for subsequent efficiency correction and torque distribution calculation.

[0050] The method includes the following steps:

[0051] S1: Obtain road and traffic information for the vehicle's preset route ahead.

[0052] The vehicle navigation system uses electronic maps to obtain information on road slope and curvature.

[0053] Real-time traffic information is obtained from road infrastructure and traffic management systems through vehicle-to-everything (V2X) communication. This real-time traffic information includes traffic light status, phase timing, traffic flow congestion status, and speed limit information.

[0054] It should be noted that the preset itinerary is preferably generated in real time by the vehicle's advanced driver assistance system domain controller or vehicle controller based on the current navigation path planning results. The controller takes the current vehicle position as the starting point and extracts a segment of the planned path. The length of this path can be dynamically configured based on the current vehicle speed and the main controller's computation cycle to cover, for example, the expected travel range of the next 30 seconds to 2 minutes. Specifically, based on the current vehicle speed and combined with a preset prediction time window, integral calculations can be performed on the navigation path curve to dynamically delineate a trajectory line with a specific sequence of latitude and longitude points on the map. This trajectory line serves as the spatial reference for subsequent road and traffic information acquisition and prediction optimization calculations.

[0055] The electronic map is preferably a high-precision map database with lane-level alignment and attribute data. The "slope information" can be the longitudinal elevation change rate or slope angle value sampled along the lane centerline at preset spatial intervals (e.g., every 1-5 meters), used to characterize the road's longitudinal slope; the "curvature information" can be the road turning radius or curvature value at the corresponding sampling point, used to characterize the road's planar geometry. The vehicle uses a GNSS positioning module, preferably combined with data from an inertial measurement unit, for fusion positioning to determine its precise position within the lane. The multi-source information fusion processing unit calls the high-precision map data via an onboard communication bus (e.g., onboard Ethernet or CANFD bus) and runs a map matching algorithm to match the vehicle trajectory points obtained from real-time positioning with the lane lines in the high-precision map, thereby indexing and extracting the slope value sequence and curvature value sequence corresponding to the preset travel trajectory line from the map database. The sequence is further organized into a data array with time or spatial markers for subsequent use in the vehicle longitudinal dynamics prediction model.

[0056] Real-time traffic information can be acquired collaboratively through both cellular vehicle-to-everything (V2X) communication and short-range direct communication. For traffic light information, after a vehicle enters a certain range of an intersection, the onboard wireless communication unit can receive structured signal phase and time (SPAT) messages periodically broadcast by roadside units via direct communication methods such as C-V2X or DSRC. The communication protocol stack in the multi-source information fusion processing unit decodes these SPAT messages, parsing the color status of traffic lights in each direction at the intersection ahead, the remaining time of the current phase, and the forecast information for the next phase, thereby obtaining precise passage time window constraints. For traffic congestion status and speed limit information, the vehicle's connected communication module or T-BOX can interact with the traffic information cloud platform via the cellular network, reporting its location and preset travel range to the cloud platform. Based on floating car data and road network status assessment algorithms, the cloud platform generates the congestion level of the target road segment, the predicted average speed, and the officially released static or dynamic speed limit values, and sends them to the vehicle in a structured data format (e.g., JSON or binary protocol). The vehicle-mounted multi-source information fusion processing unit analyzes the data to obtain the congestion status and speed limit information corresponding to each road segment of the preset route.

[0057] Because the sources of slope information, curvature information, traffic light status and phase sequence, congestion status, and speed limit information are different, their update intervals are different, and they have their own timestamps, a data cleaning and fusion module can also run in the multi-source information fusion processing unit to unify the above multi-source data under the time reference of the vehicle controller. Specifically, a unified time axis based on the vehicle controller clock can be established, and various types of data can be resampled according to the controller's main control cycle (e.g., 10-50 milliseconds): for quantities with long update cycles (such as slope sequences updated every 1 second, congestion indices updated every several seconds), the previous valid value can be preserved and / or linear interpolation can be used to complete them between adjacent control cycles; for quantities with high update frequencies (such as traffic light phase data broadcast once every 100 milliseconds), interpolation or alignment can be performed according to the timestamp. Through the above alignment and fusion processing, a comprehensive road condition information table that changes with time and / or space is formed within the preset travel range. The comprehensive road condition information table provides corresponding slope, curvature, traffic light constraints, congestion status and speed limit information for each prediction node, providing a consistent, complete and calculable environmental input for subsequent demand power prediction and torque planning steps.

[0058] S2: Based on the road and traffic information, predict the vehicle power demand during the future driving cycle, and generate a total demand torque command sequence accordingly.

[0059] Based on road slope, road curvature and road speed limit information, a longitudinal dynamics model of the vehicle is established in a preset prediction time domain, and the traffic light status and phase sequence are used as vehicle speed constraints.

[0060] With the goal of minimizing the vehicle's driving energy consumption in the predicted time domain, a model predictive control algorithm is used to perform rolling optimization on the vehicle's longitudinal dynamics model to obtain a reference vehicle speed sequence for future driving cycles.

[0061] Based on the reference vehicle speed sequence, road gradient information, and the vehicle longitudinal dynamics model, the total demand torque of the vehicle at each moment required to follow the reference vehicle speed sequence is calculated, and the total demand torque command sequence is formed from the total demand torque of the vehicle at each moment.

[0062] It should be noted that the vehicle longitudinal dynamics model preferably adopts a discretized longitudinal motion equation form to describe the evolution of vehicle speed over time under a given driving or braking action. At each prediction step (e.g., 0.1–0.5 seconds), the model comprehensively considers the vehicle mass and its generated inertial force, the gravity component along the road surface caused by the road slope, the air resistance determined by the air resistance coefficient and the vehicle's frontal area, and the rolling resistance determined by the rolling resistance coefficient. Based on the current or predicted vehicle speed, acceleration, and the road slope value at the corresponding position obtained in step S1, it calculates the vehicle speed and position at the next discrete moment. The preset prediction time domain can correspond to a preset travel distance, for example, covering several tens of seconds of travel time in the future, and is discretized into several prediction steps to form a state sequence within the prediction time domain. Regarding constraint injection, road speed limit information serves as a hard upper limit for vehicle speed; road curvature information can be combined with lateral acceleration limits to derive the safe passing speed upper limit at each prediction node, thereby forming additional speed upper bound constraints. Traffic signal light states and phase sequences are transformed into time-space related passage constraints: the controller calculates whether a vehicle can pass through the intersection within the green light window under different reference speed trajectories based on the current vehicle position, predicted speed, and remaining red and green light phases; if it can, a speed range constraint of "reaching the intersection within the green light time window" is applied; if it cannot, a speed constraint of "decelerating to zero before the stop line during the red light" is applied. Traffic flow congestion can be modeled as soft speed limits or expected following distance constraints for each road segment, achieved by adding penalty terms to the optimization objective or constraints.

[0063] Furthermore, after establishing the longitudinal dynamics model and related constraints, an energy consumption optimization problem can be constructed within a preset prediction time domain. Specifically, the instantaneous driving power at each discrete moment within the prediction time domain is weighted and accumulated to obtain an approximate expression for the vehicle's driving energy consumption. The instantaneous driving power can be calculated by combining the product of vehicle speed and total driving torque with transmission efficiency. The optimization variables can be selected as acceleration commands, driving force commands, or reference vehicle speed sequences at each prediction moment. With minimizing the vehicle's driving energy consumption within the prediction time domain as the optimization objective, and under the premise of satisfying longitudinal dynamics constraints, vehicle speed upper and lower bound constraints, and traffic light and congestion constraints, the model predictive control algorithm is used to solve this optimization problem. In implementation, the continuous problem can be discretized and transformed into a quadratic programming or nonlinear programming problem, which can be solved using the interior-point method, the effective set method, or a pre-configured numerical solver. At the beginning of each control cycle (e.g., 20–50 milliseconds), the controller acquires the current actual vehicle state and the latest road and traffic information, solves for a set of optimal control quantity sequences and corresponding reference vehicle speed sequences in the future prediction time domain, and executes only the control results of the most recent one or a few moments. In the next control cycle, the optimization problem is reconstructed and solved based on the updated state and environmental information, thereby achieving rolling forward closed-loop optimization. The reference vehicle speed sequence obtained in the above manner is not only kinetically achievable, but also, considering factors such as gradient, speed limit, traffic lights, and congestion, makes the overall vehicle driving energy consumption close to optimal.

[0064] Furthermore, after obtaining the reference vehicle speed sequence for the future driving cycle, the predictive energy management controller performs inverse dynamics calculations based on the vehicle's longitudinal dynamics model. Specifically, by performing a differential operation on the reference vehicle speed sequence, a reference acceleration sequence for each discrete moment is obtained; combined with the road slope information corresponding to each moment, this information is substituted into the longitudinal mechanics equations containing inertial force, slope gravity components, air resistance, and rolling resistance to calculate the total longitudinal resistance that the vehicle needs to overcome at each predicted moment to track the reference vehicle speed trajectory, thereby obtaining the corresponding total required driving force. Then, using the dynamic radius of the vehicle wheels and the gear ratio of the final drive, the total required driving force is converted into the total required torque at the output shaft of the electric drive assembly. The above calculation is repeated at all discrete moments within the prediction time domain to form a total required torque command sequence that corresponds one-to-one with the reference vehicle speed sequence in time. This total demand torque command sequence can be used to describe the power demand distribution of the vehicle during the future driving cycle, and also serves as the upper-level input for torque distribution and motor control command synthesis in the subsequent electric drive assembly energy consumption optimization control, realizing a complete technical link connection from the road and traffic information ahead to the electric drive assembly torque control.

[0065] S3: During the current operation of the vehicle, the permanent magnet synchronous motor is identified online in real time to obtain the real-time electromagnetic parameters of the permanent magnet synchronous motor.

[0066] During the drive control cycle of the permanent magnet synchronous motor, a preset high-frequency voltage signal is superimposed on the fundamental voltage vector output by the inverter to excite the generation of a high-frequency response current containing motor inductance information.

[0067] The three-phase current of the synchronously sampled motor is separated from the three-phase current by coordinate transformation and bandpass filtering to extract the current response component that is in the same frequency as the high-frequency voltage signal.

[0068] Based on the amplitude and frequency of the high-frequency voltage signal and the separated current response components, the direct-axis inductance and quadrature-axis inductance of the permanent magnet synchronous motor are calculated in real time using the high-frequency impedance model of the permanent magnet synchronous motor, and are used as the real-time electromagnetic parameters.

[0069] It should be noted that the preset high-frequency voltage signal can be generated by a digital signal processor (DSP) or microcontroller unit (MCU) within the controller. It is a sinusoidal voltage vector with constant amplitude and a frequency much higher than the motor's fundamental operating frequency. Its frequency is typically selected in the range of several hundred hertz to several thousand hertz to avoid the motor's fundamental frequency and major harmonics, and is lower than the inverter's PWM carrier frequency, thereby avoiding aliasing interference. The amplitude of this high-frequency signal is pre-calibrated to be small enough that the resulting additional high-frequency torque ripple has a negligible impact on driving smoothness, and large enough to ensure a discernible current response in the context of measurement noise. Within each drive control cycle (e.g., a PWM cycle), the high-frequency voltage vector is synthesized in the digital domain with the fundamental voltage vector used to generate the fundamental electromagnetic torque by modifying the inverter's space vector pulse width modulation duty cycle. This synthesized signal is then output to the motor's stator three-phase windings by the inverter's power switching devices, achieving uninterrupted high-frequency excitation of the motor.

[0070] Furthermore, to capture the motor's response to the high-frequency excitation, the controller synchronously acquires the motor's three-phase stator current at a sampling frequency several times higher than the injected signal using a stator current sensor and a high-precision analog-to-digital converter. The acquired three-phase current is first transformed from the three-phase stationary coordinate system to a d / q rotating coordinate system based on the rotor flux linkage through Clarke and Parker transforms. In this coordinate system, the fundamental component of the current behaves as an approximately DC quantity in steady state, while the superimposed high-frequency voltage excitation corresponds to an AC current component oscillating at the injected frequency. Utilizing this spectral characteristic, the online parameter identification module applies a bandpass filter or an equivalent synchronous demodulation algorithm with its center frequency locked to the injected frequency to the current signal in the d / q coordinate system, filtering out the fundamental component, switching harmonics, and other noise, retaining only the high-frequency current response component with the same frequency as the injected signal, thereby obtaining the high-frequency current vector for parameter identification.

[0071] Under high-frequency excitation conditions, the back electromotive force of a permanent magnet synchronous motor (PMSM) has a relatively small impact, and the voltage-current relationship at the motor ports can be described by a high-frequency equivalent impedance model dominated by inductance. The online parameter identification module uses the known amplitude and frequency of the high-frequency injected voltage vector, along with the extracted high-frequency current response vector, to calculate the equivalent high-frequency impedance in the direct-axis and quadrature-axis directions in the d / q coordinate system. Combining the relationship between impedance and inductance in the high-frequency impedance model, the direct-axis and quadrature-axis inductances of the PMSM under the current operating conditions can be calculated in real time through algebraic operations. This calculation process can be repeated in each drive control cycle or every few control cycles, allowing the obtained inductance parameters to track the electromagnetic characteristic drift caused by temperature changes, changes in magnetic saturation, and material aging.

[0072] Furthermore, to improve the stability of parameter identification results, the controller can perform moving average or first-order low-pass filtering on the continuously calculated direct-axis and quadrature-axis inductances to suppress instantaneous measurement noise and high-frequency fluctuations. Simultaneously, operating condition judgment logic can be set to maintain the previously reliable parameter values ​​under conditions such as excessively low motor current, excessively high speed, or insufficient signal-to-noise ratio of the injected signal, avoiding adverse effects of abnormal data on subsequent control. The processed direct-axis and quadrature-axis inductances are stored as real-time electromagnetic parameters in the parameter area of ​​the motor controller and used in subsequent steps to dynamically correct the reference efficiency mapping curve of the permanent magnet synchronous motor. This ensures that the torque control strategy based on efficiency mapping remains consistent with the current true electromagnetic state of the motor, thereby providing a reliable underlying parameter foundation for overall energy consumption optimization.

[0073] S4: Dynamically correct the reference efficiency mapping curve based on the real-time electromagnetic parameters.

[0074] Obtain the baseline efficiency mapping curve and its associated motor parameterized mathematical model, wherein the motor parameterized mathematical model defines the mapping relationship between the motor electromagnetic parameters and the optimal efficiency operating point of the permanent magnet synchronous motor.

[0075] Substitute the direct-axis inductance and quadrature-axis inductance from the real-time electromagnetic parameters into the parameterized mathematical model of the motor, and update the model parameters in the mapping relationship;

[0076] Based on the updated mapping relationship, the baseline efficiency mapping curve is recalculated or interpolated to generate a corrected efficiency mapping curve that reflects the actual characteristics of the current motor.

[0077] It should be noted that the reference efficiency mapping curve is preferably stored in the form of a two-dimensional lookup table, with the horizontal axis representing motor speed and the vertical axis representing output torque. Each table cell stores the reference efficiency data under the corresponding speed and torque conditions, and / or the combination of control parameters used to achieve high-efficiency operation, such as the optimal ratio of direct-axis current components to quadrature-axis current components, or the target point on the maximum torque per ampere control trajectory. The lookup data is usually derived from the factory test calibration of the motor under standard temperature and cooling conditions. The "parametric mathematical model of the motor" associated with this lookup table is used to describe the correspondence between motor losses such as copper and iron losses, as well as overall efficiency, and motor electromagnetic parameters. The motor electromagnetic parameters include at least the direct-axis inductance, quadrature-axis inductance, and flux linkage corresponding to the rotor permanent magnets. The control variables include the direct-axis current component and the quadrature-axis current component. This model can be a simplified loss model based on physical mechanisms, or an empirical model obtained by fitting experimental data, used to calculate the losses and efficiency at the corresponding operating point given inductance parameters and current commands.

[0078] Furthermore, the efficiency curve dynamic management module reads the real-time direct-axis inductance and real-time quadrature-axis inductance, smoothed in step S3, from the shared storage area and substitutes them as the latest electromagnetic parameters into the aforementioned parameterized mathematical model, replacing the nominal inductance parameters previously used in the model or the inductance parameters from the last update. Through this update process, the physical coefficients related to inductance within the model are recalculated or proportionally corrected, allowing the flux linkage distribution, magnetic saturation, and iron loss characteristics calculated by the model under different torque and speed conditions to more closely approximate the actual state of the current motor. For example, in the control trajectory model used to determine the optimal ratio of the direct-axis current component and the quadrature-axis current component, the real-time direct-axis inductance and quadrature-axis inductance directly affect the direction of the optimal current vector; in the efficiency model based on loss minimization, the real-time inductance is used to correct the calculation of stator flux linkage and core losses. Through the above online reparameterization process, the motor is mapped from its "factory calibration state" to its "equivalent motor under the current operating state" at the software level.

[0079] Furthermore, after updating the parameterized mathematical model, the efficiency curve dynamic management module reconstructs the baseline efficiency mapping curve based on the updated model. To balance real-time performance and computational load, one of two strategies or a combination thereof can be employed:

[0080] Recalculation Strategy: When the controller has sufficient computing power, the module can recalculate all or part of the speed-torque operating points covered by the efficiency mapping curve in the background. For each operating point, the updated mathematical model is called to solve for the high-efficiency control parameters and efficiency values ​​corresponding to that operating point under the current direct-axis inductance and quadrature-axis inductance conditions, and the new results are used to overwrite the original lookup table data. The resulting new graph comprehensively reflects the efficiency distribution under the current electromagnetic parameters.

[0081] Interpolation synthesis or local update strategy: To further reduce real-time computation, the controller can pre-store multiple sets of "basic efficiency curve clusters" based on different combinations of inductance parameters during the factory or calibration phase. When real-time inductance parameters are obtained during operation, the module compares them with the inductance parameters corresponding to each basic curve. Through multi-dimensional interpolation, weighted averaging, or similarity matching, it synthesizes an efficiency mapping curve that most closely matches the current inductance parameters from these basic curves and loads it as the current operating curve. Simultaneously, for the current and likely near-term operating speed and torque ranges of the vehicle, local fine-tuning can be performed on this basis to improve the accuracy of critical areas.

[0082] The "corrected efficiency mapping curve" generated by any of the above strategies numerically reflects the impact of real-time changes in direct-axis and quadrature-axis inductance on the optimal efficiency working state at each operating point, forming a dynamic efficiency spectrum that matches the current electromagnetic state of the motor.

[0083] The generated corrected efficiency mapping curve can be stored in an independent data area or updated incrementally on the original baseline efficiency mapping curve, and can include auxiliary information such as current electromagnetic parameters and update time. To avoid control strategy jitter caused by frequent updates, the controller can perform moving average or low-pass filtering smoothing on the efficiency values ​​and control parameters calculated multiple times at the same operating point. Finally, the corrected efficiency mapping curve is provided to the subsequent torque distribution and current command generation modules: when subsequent steps are at a given speed and torque operating point, the corresponding high-efficiency control parameters will be queried based on the corrected efficiency mapping curve, instead of directly using the factory-calibrated static efficiency data. Through this dynamic correction mechanism, even under conditions of long-term motor operation, significant temperature rise, or gradual magnetic circuit saturation, the energy consumption optimization control based on efficiency mapping in this invention can still closely follow the true efficiency characteristics of the motor, effectively supporting the continuous optimization of vehicle energy consumption.

[0084] S5: Based on the corrected baseline efficiency mapping curve, the total demand torque command sequence is decomposed and synthesized into real-time control commands for each permanent magnet synchronous motor.

[0085] Based on the total torque demand command at the current moment, the torque allocation command for each permanent magnet synchronous motor is determined according to the principle of optimizing the real-time comprehensive efficiency of the electric drive assembly based on the modified efficiency mapping curve.

[0086] Based on the torque distribution command and current speed of each permanent magnet synchronous motor, the corresponding corrected efficiency mapping curve is queried to obtain the optimal efficiency current command of each permanent magnet synchronous motor under the corresponding operating conditions.

[0087] The optimal efficiency current command is used as the real-time control command.

[0088] It should be noted that the so-called "allocation principle based on the optimal real-time comprehensive efficiency of the electric drive assembly" means minimizing the sum of losses of multiple motors under the current operating conditions while meeting the current total torque demand constraint. In specific implementation, the multi-motor collaborative control unit first reads the current total torque demand value from the total torque demand command sequence and obtains the current speed, allowable output torque range, thermal state, and vehicle stability control-related constraints of each permanent magnet synchronous motor. Subsequently, based on the corrected efficiency mapping curves of each motor, the control unit constructs several candidate torque allocation schemes that meet the condition that "the sum of the output torques of each motor equals the total torque demand." For each candidate scheme, the control unit searches for the efficiency or equivalent loss value at the current speed and candidate torque operating point in the corrected efficiency mapping curves corresponding to each motor, and then sums the losses of each motor to obtain the instantaneous total loss of the electric drive assembly under that candidate scheme. By comparing the candidate schemes, the torque allocation result with the lowest instantaneous total loss or the highest comprehensive efficiency is selected as the torque allocation command for each permanent magnet synchronous motor in this control cycle. To avoid drastic torque fluctuations between adjacent control cycles, the control unit can also impose constraints or penalties on the rate of torque change during optimization to balance efficiency and drive smoothness.

[0089] Furthermore, after obtaining the torque distribution commands from each permanent magnet synchronous motor, the multi-motor collaborative control unit needs to convert the torque target into a current target value that the motor controller can directly execute. To this end, the control unit first synchronously acquires the real-time speed of each motor, combining the "torque distribution command" and the "current speed" into specific operating point coordinates. Then, using this operating point as an index, a lookup table or two-dimensional interpolation is performed in the corresponding motor's corrected efficiency mapping curve lookup table to obtain a set of current control parameters that maximize the motor's operating efficiency under the given speed and torque conditions. These current control parameters can be understood as the "optimal efficiency current command" under the current operating conditions, satisfying both the required torque output and minimizing the motor's own copper losses, iron losses, and other comprehensive losses. In this way, this step not only completes the distribution of total torque among multiple motors but also ensures that the torque allocated to each motor is generated in the most energy-efficient manner for that motor at that time.

[0090] Furthermore, the multi-motor cooperative control unit standardizes and encapsulates the optimal efficiency current commands of each permanent magnet synchronous motor, forming real-time control command data in a unified format. The control commands include at least the target motor's identification information, the corresponding current setpoint, and a timestamp. The control unit synchronously sends these real-time control commands to their respective motor controllers within each control cycle via a high-speed vehicle communication network (e.g., CANFD or vehicle Ethernet). Upon receiving the real-time control commands, each motor controller directly uses the current setpoint as the target setpoint for current closed-loop control. This target setpoint will be used by each motor controller in subsequent steps (S6) to generate specific voltage commands and switching signals, achieving precise output of the motor's electromagnetic torque. Through the above process, the energy-saving driving strategy obtained in step S2 based on the road condition planning ahead, after adaptive correction of motor state and efficiency in steps S3 and S4, is finally implemented in this step as the optimal current control action of each motor, thereby achieving optimal collaborative control of the electric drive assembly's energy consumption while ensuring the vehicle's overall power requirements.

[0091] S6: Execute the real-time control command to drive the permanent magnet synchronous motor to run.

[0092] Based on the current command in the real-time control command, a voltage command is generated in the current loop controller;

[0093] Based on the current DC bus voltage, real-time speed and voltage command of each permanent magnet synchronous motor, the pulse width modulation strategy of the inverter is dynamically adjusted.

[0094] Based on the pulse width modulation strategy, a corresponding switching signal for the power switching device is generated, which drives the power switching device of the inverter to operate, so that each permanent magnet synchronous motor outputs a drive torque corresponding to the torque distribution command.

[0095] It should be noted that after receiving real-time control commands, each motor controller extracts the optimal efficiency current command and sets it as the setpoint for the current loop controller. Simultaneously, the controller collects the three-phase stator current of the motor through current sensors and obtains the actual current components in the rotating coordinate system through coordinate transformation. The current loop controller compares the current setpoint with the actual current value, calculates the current error, and uses algorithms such as proportional-integral regulation to generate corresponding voltage commands in each control cycle to eliminate this error. Through this rapid closed-loop regulation process, the motor stator current can closely track the optimal efficiency current command, thereby establishing an electromagnetic state within the motor that matches the torque distribution command, providing a precise electromagnetic basis for efficient torque output.

[0096] Furthermore, the generated voltage command needs to be implemented through the switching action of the inverter's power switching devices. To achieve the required voltage vector while minimizing the inverter's own losses, a dynamic adjustment module for pulse width modulation (PWM) strategy is installed in the motor controller. This module acquires the DC bus voltage and motor speed in real time, and selects or adjusts the inverter's PWM mode based on the amplitude and phase of the current voltage command. Specifically, under low-speed, low-torque, and low-modulation-ratio conditions, a modulation mode that reduces the number of switching operations on certain bridge arms within a switching cycle can be preferred to reduce switching losses. Under light-load conditions with low electromagnetic noise requirements, the carrier frequency of the PWM can be appropriately reduced to further reduce switching losses. In the high-modulation region approaching the inverter's linear output limit, waveform distortion is reduced and additional losses are avoided by optimizing the voltage vector trajectory and modulation mode. Through the above dynamic optimization, the voltage command is reasonably mapped to a PWM mode that the inverter can execute under the current bus voltage and speed conditions, balancing motor control accuracy and the energy efficiency of the electric drive system itself.

[0097] Furthermore, after determining the specific pulse width modulation strategy, the digital pulse generation unit inside the motor controller calculates the on-time and off-time of each power switching device within the current control cycle based on the selected modulation mode and carrier parameters, and generates a switching control signal with a corresponding duty cycle and phase relationship. To avoid two switching devices on the same bridge arm conducting simultaneously, the controller introduces a preset dead time before outputting the switching signal and performs power amplification and electrical isolation of the signal through an isolation drive circuit. Finally, the power switching devices in the inverter perform high-speed on- and off-state operations according to the aforementioned switching control signal, so that the DC bus voltage is synthesized into a three-phase AC voltage and current consistent with the aforementioned voltage command on the three-phase windings of the motor stator.

[0098] Under the closed-loop control of the current loop, the stator current stably follows the changes in the optimal efficiency current command. The permanent magnet synchronous motor thus generates electromagnetic torque corresponding to the torque distribution command in step S5, and outputs it to the wheels through a reduction gear or direct drive. At this point, the complete technical link between the upper-level energy consumption optimization decision based on the road conditions ahead and real-time motor parameters, and the lower-level inverter switching action and motor torque output, is finally closed in this step, realizing the physical implementation of the new energy vehicle electric drive assembly energy consumption optimization method described in this invention.

[0099] Example 2 is the second embodiment of the present invention, which differs from the previous embodiment in that:

[0100] If the aforementioned functions are implemented as software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this invention, or the part that contributes to the prior art or the current technical solution, can be embodied in the form of a software product. This current computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of this invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0101] The logic and / or steps represented in the flowchart or otherwise described herein, for example, can be considered as a sequenced list of executable instructions for implementing logical functions, and can be embodied in any computer-readable medium for use by, or in conjunction with, an instruction execution system, apparatus, or device (such as a computer-based system, a processor-including system, or other system that can fetch and execute instructions from, an instruction execution system, apparatus, or device). For the purposes of this specification, "computer-readable medium" can be any means that can contain, store, communicate, propagate, or transmit programs for use by, or in conjunction with, an instruction execution system, apparatus, or device.

[0102] More specific examples of computer-readable media (a non-exhaustive list) include: electrical connections (electronic devices) having one or more wires, portable computer disk drives (magnetic devices), random access memory (RAM), read-only memory (ROM), erasable and editable read-only memory (EPROM or flash memory), fiber optic devices, and portable optical disc read-only memory (CDROM). Furthermore, computer-readable media can even be paper or other suitable media on which the program can be printed, because the program can be obtained electronically, for example, by optically scanning the paper or other medium, followed by editing, interpreting, or otherwise processing as necessary, and then stored in computer memory.

[0103] Example 3, an embodiment of the present invention, provides an energy consumption optimization system for electric drive assembly of new energy vehicles, including an information acquisition module, a torque planning module, a parameter identification module, a curve correction module, and an instruction synthesis module.

[0104] Information acquisition module: Acquires road and traffic information for the vehicle's preset route ahead;

[0105] Torque planning module: Based on the road and traffic information, predicts the vehicle's power demand during future driving cycles and generates a total demand torque command sequence accordingly;

[0106] Parameter identification module: During the current operation of the vehicle, the permanent magnet synchronous motor is identified online in real time to obtain the real-time electromagnetic parameters of the permanent magnet synchronous motor;

[0107] Curve correction module: Dynamically corrects the reference efficiency mapping curve based on the real-time electromagnetic parameters;

[0108] Command synthesis module: Based on the corrected benchmark efficiency mapping curve, the total demand torque command sequence is decomposed and synthesized into real-time control commands for each permanent magnet synchronous motor.

[0109] Drive control module: Executes the real-time control commands to drive the permanent magnet synchronous motor.

[0110] Example 4 is an embodiment of the present invention, which provides a method for optimizing the energy consumption of electric drive assembly in new energy vehicles. In order to verify the beneficial effects of the present invention, scientific demonstration is carried out through economic benefit calculation and simulation / comparative experiments.

[0111] This embodiment verifies the effectiveness of the predictive energy management method based on online parameter identification and dynamic efficiency curve correction described in this invention in optimizing the energy consumption of the dual-motor electric drive assembly under real-world complex road conditions through a real-vehicle road test. The test vehicle is a dual-motor four-wheel drive pure electric SUV, with one permanent magnet synchronous motor with a rated power of approximately 150kW installed on each of the front and rear axles. The electric drive domain controller has the ability to switch between multiple software strategies online, and the hardware configuration remains consistent across all schemes.

[0112] To enhance the test's sensitivity to the time-varying characteristics of motor parameters, the front axle motor underwent prolonged high-load bench preheating before the test, stabilizing its winding temperature at approximately 85°C, significantly higher than the efficiency chart calibration temperature. The rear axle motor underwent only routine preheating, with its stable temperature around 30°C. This allows for the creation of a differential operating condition—the front motor undergoing thermal degradation while the rear motor approaches its calibration state—without hardware replacement, to verify the practical effectiveness of online parameter identification and dynamic efficiency mapping correction.

[0113] The road test route was a closed loop with a total length of approximately 25.6 km. Approximately 40% of the route consisted of low-speed urban congestion (average speed approximately 22 km / h, including 8 traffic light intersections), approximately 30% was urban expressway (average speed approximately 65 km / h), and approximately 30% was highway (average speed approximately 100 km / h). During the test, the ambient temperature was maintained at around 25℃, and operating parameters such as tire pressure, vehicle weight, and air conditioning settings remained consistent across all tests. The initial state of charge of the power battery was controlled at 80% ± 2%.

[0114] To compare and evaluate the effectiveness of the present invention, this embodiment sets up two control strategies:

[0115] Option C0 (Control Group): The conventional torque demand calculation based on pedal opening and vehicle speed is adopted, and an energy management strategy with simple vehicle speed prediction capability is introduced. The static efficiency mapping curve calibrated by the factory is used for torque distribution and current command generation. The torque distribution of multiple motors is mainly based on the calibration efficiency diagram and empirical rules, without online electromagnetic parameter identification and dynamic correction of efficiency curves.

[0116] Solution C1 (Method of the Invention): First, high-precision electronic maps and vehicle-to-everything (V2X) communication are used to obtain information such as road slope, curvature, traffic light phase, congestion status, and speed limits for the preset route ahead. Then, within a prediction time domain of approximately 2.5 km, a rolling reference vehicle speed is planned with the goal of minimizing the energy consumption of the electric drive system, generating a sequence of total demand torque commands. During vehicle operation, the motor controller identifies the current electromagnetic parameters such as the direct-axis and quadrature-axis inductances of the front and rear motors online through high-frequency voltage signal injection and response current analysis, and dynamically corrects the original baseline efficiency mapping curve based on these real-time parameters. Finally, the multi-motor collaborative control unit allocates the total demand torque between the front and rear motors based on the corrected efficiency mapping curve, and generates corresponding optimal efficiency current commands, which are then sent to each motor controller for execution.

[0117] During actual testing, road and traffic information ahead was updated and input into the predictive controller at a cycle of approximately 50 milliseconds. The predictive controller performed rolling optimization at a cycle of approximately 100 milliseconds, outputting the reference vehicle speed and the corresponding total torque demand. The online parameter identification module corrected the equivalent inductance of the front and rear motors in real time at intervals of several motor control cycles, and periodically updated the key grid points in the efficiency mapping curve related to the current operating conditions. Each control strategy was tested at least three times on the same route. After removing samples with abnormal traffic disturbances, the average value of the remaining data was taken as the result. Please refer to Table 1 for specific data.

[0118] Table 1: Experimental Results Data Table

[0119]

[0120] The energy-saving effect of the method of this invention can be intuitively judged from the two core indicators of "total energy consumption" and "equivalent driving range". Scheme C0 has a total energy consumption of 6.18 kWh under this mixed operating condition, while Scheme C1 reduces it to 5.81 kWh, a reduction of approximately 5.99%. With the same available battery energy, the equivalent driving range increases from 312.4 km to 331.8 km, an improvement of approximately 6.21%. Given that the electric drive system itself is already at a high efficiency level, this level of improvement has significant engineering implications, indicating that the predictive performance optimization and online model correction proposed in this invention are not minor adjustments, but rather bring about a considerable improvement in overall vehicle energy consumption.

[0121] Further analysis of the "average efficiency of the electric drive assembly" and the average efficiency data of the front and rear motors reveals that the technical advantages of this invention stem from two aspects: "model approximation to reality" and "more reasonable allocation." The average efficiency of the assembly increased from 92.8% to 93.7%, an increase of approximately 0.97 percentage points; among which, the average efficiency of the front motor increased from 91.7% to 93.2%, an increase of approximately 1.64 percentage points, while the efficiency of the rear motor remained at a relatively high level, with only a slight increase. Test conditions show that the front motor, due to hot operation, deviates more significantly from its calibrated electromagnetic parameters. If static efficiency mapping is still used for torque distribution and current control, the motor will operate in the "theoretically optimal but actually deviated" region for an extended period. In scheme C0, even with some forward-looking speed planning, the torque distribution and current commands are still based on the old model, resulting in a significantly lower efficiency for the front motor.

[0122] In contrast, this invention obtains the real-time electromagnetic parameters of the front motor at high temperatures through online parameter identification and dynamically corrects the efficiency mapping curve accordingly, enabling the efficiency graph to reflect its current true loss characteristics. Based on this, the multi-motor torque distribution algorithm actively allocates more torque to the rear motor, which maintains higher efficiency, while appropriately "avoiding" the target operating conditions of the front motor, ensuring that the front motor operates within its local high-efficiency zone under its current state. This closed-loop mechanism of "identification, correction, and decision-making" allows the system to maintain near-optimal efficiency distribution even with significant drift in motor parameters, achieving a dynamic adaptive control effect that traditional static calibration methods cannot achieve.

[0123] The two indicators, "peak temperature rise of the front motor" and "average switching loss of the inverter," reflect the additional benefits of this invention in terms of reliability and device losses. The peak temperature rise of the front motor decreased from 58.2℃ to 50.1℃, a reduction of nearly 14%. Lower temperature rise means a significant reduction in thermal stress on the stator windings and insulation system, which helps extend motor life and slows down the further drift of electromagnetic parameters over time. The average switching loss of the inverter decreased from 225W to 198W, a reduction of approximately 12%. This improvement did not come from hardware changes, but rather from the adaptive adjustment of the pulse width modulation strategy at the execution layer: appropriately reducing the switching frequency or using a modulation mode with fewer switching operations during low-speed, low-load conditions, and switching back to a high-frequency continuous modulation mode when high dynamic response is required, thereby reducing the dynamic losses of devices while ensuring control performance. This fine-grained optimization of switching losses is usually ignored in traditional methods, but this invention incorporates it into the overall energy consumption optimization framework, reflecting a comprehensive trade-off approach from a system perspective.

[0124] The increase in "energy recovery" also confirms the synergistic effect of forward-looking torque planning and dynamic model correction in regenerative braking. Energy recovery increased from 0.71 kWh to 0.79 kWh, an increase of approximately 11.27%. In the method of this invention, the predictive controller can use information such as traffic light phase and gradient to plan the deceleration process in advance, and combine this with a real-time corrected efficiency mapping curve to select a more favorable regenerative braking configuration for both motors. This allows regenerative braking to operate in the most efficient range as much as possible without affecting comfort, a fact verified by the quantified increase in recovered energy.

[0125] It should be noted that the method of this invention does introduce certain computational overhead, with the "computational power ratio of prediction and identification algorithms" increasing from 9.6% to 18.8%. However, this ratio is still within the acceptable range of current mainstream automotive-grade domain controllers, and compared with the energy-saving benefits and lifespan benefits brought by reduced thermal load, this computational cost is a reasonable engineering trade-off. More importantly, the additional computational power is mainly used for online identification, motor model correction, and optimization solutions. This investment directly improves the realism of the control model and the refinement of decision-making, which is the key support for the "soft upgrade" achieved by this invention compared to traditional methods.

[0126] In summary, this embodiment, through comparative test data under real road mixed conditions, fully demonstrates that the "predictive energy management method based on online parameter identification and dynamic curve correction" proposed in this invention has significant energy-saving effects and engineering feasibility. Under the same hardware platform and similar computing resource constraints, this invention can simultaneously achieve improvements in multiple dimensions such as energy consumption, efficiency, temperature rise, and device losses, demonstrating creativity and substantial progress compared to existing technologies.

[0127] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention, and all such modifications or substitutions should be covered within the scope of the claims of the present invention.

Claims

1. A method for optimizing the energy consumption of an electric drive assembly in a new energy vehicle, wherein the electric drive assembly includes at least two permanent magnet synchronous motors and their controllers, and the controllers are pre-set with a reference efficiency mapping curve for the permanent magnet synchronous motors, characterized in that, The method includes the following steps: Obtain road and traffic information for the vehicle's preset route ahead; Based on the road and traffic information, predict the vehicle power demand during future driving cycles and generate a total demand torque command sequence accordingly. During the current operation of the vehicle, the permanent magnet synchronous motor is identified online in real time to obtain the real-time electromagnetic parameters of the permanent magnet synchronous motor. The reference efficiency mapping curve is dynamically corrected based on the real-time electromagnetic parameters. Based on the revised baseline efficiency mapping curve, the total demand torque command sequence is decomposed and synthesized into real-time control commands for each permanent magnet synchronous motor. The real-time control command is executed to drive the permanent magnet synchronous motor to run; The dynamic correction of the baseline efficiency mapping curve includes: obtaining the baseline efficiency mapping curve and its associated motor parameterized mathematical model, wherein the motor parameterized mathematical model defines the mapping relationship between the electromagnetic parameters of the motor and the optimal efficiency operating point of the permanent magnet synchronous motor; substituting the direct-axis inductance and quadrature-axis inductance in the real-time electromagnetic parameters into the motor parameterized mathematical model to update the model parameters in the mapping relationship; and based on the updated mapping relationship, recalculating or interpolating the baseline efficiency mapping curve to generate a corrected efficiency mapping curve that reflects the actual characteristics of the current motor. The step of decomposing and synthesizing the total demand torque command sequence into real-time control commands for each permanent magnet synchronous motor includes: determining the torque allocation command for each permanent magnet synchronous motor based on the current total demand torque command, using the optimal real-time comprehensive efficiency of the electric drive assembly based on the modified efficiency mapping curve as the allocation principle; querying the corresponding modified efficiency mapping curve according to the torque allocation command and current speed of each permanent magnet synchronous motor to obtain the optimal efficiency current command for each permanent magnet synchronous motor under the corresponding operating condition; and using the optimal efficiency current command as the real-time control command.

2. The energy consumption optimization method for electric drive assembly of new energy vehicles as described in claim 1, characterized in that, The acquisition of road and traffic information for the vehicle's preset journey ahead includes: The vehicle navigation system uses electronic maps to obtain information on road slope and curvature. Real-time traffic information is obtained from road infrastructure and traffic management systems through vehicle-to-everything (V2X) communication. This real-time traffic information includes traffic light status, phase timing, traffic flow congestion status, and speed limit information.

3. The energy consumption optimization method for the electric drive assembly of new energy vehicles as described in claim 2, characterized in that, The sequence of commands to generate total demand torque includes: Based on road slope, road curvature and road speed limit information, a longitudinal dynamics model of the vehicle is established in a preset prediction time domain, and the traffic light status and phase sequence are used as vehicle speed constraints. With the goal of minimizing the vehicle's driving energy consumption in the predicted time domain, a model predictive control algorithm is used to perform rolling optimization on the vehicle's longitudinal dynamics model to obtain a reference vehicle speed sequence for future driving cycles. Based on the reference vehicle speed sequence, road gradient information, and the vehicle longitudinal dynamics model, the total demand torque of the vehicle at each moment required to follow the reference vehicle speed sequence is calculated, and the total demand torque command sequence is formed from the total demand torque of the vehicle at each moment.

4. The energy consumption optimization method for the electric drive assembly of new energy vehicles as described in claim 3, characterized in that, The online real-time parameter identification includes: During the drive control cycle of the permanent magnet synchronous motor, a preset high-frequency voltage signal is superimposed on the fundamental voltage vector output by the inverter to excite the generation of a high-frequency response current containing motor inductance information. The three-phase current of the synchronously sampled motor is separated from the three-phase current by coordinate transformation and bandpass filtering to extract the current response component that is in the same frequency as the high-frequency voltage signal. Based on the amplitude and frequency of the high-frequency voltage signal and the separated current response components, the direct-axis inductance and quadrature-axis inductance of the permanent magnet synchronous motor are calculated in real time using the high-frequency impedance model of the permanent magnet synchronous motor, and are used as the real-time electromagnetic parameters.

5. The energy consumption optimization method for the electric drive assembly of new energy vehicles as described in claim 4, characterized in that, The execution of the real-time control command includes: Based on the current command in the real-time control command, a voltage command is generated in the current loop controller; Based on the current DC bus voltage, real-time speed and voltage command of each permanent magnet synchronous motor, the pulse width modulation strategy of the inverter is dynamically adjusted. Based on the pulse width modulation strategy, a corresponding switching signal for the power switching device is generated, which drives the power switching device of the inverter to operate, so that each permanent magnet synchronous motor outputs a drive torque corresponding to the torque distribution command.

6. A new energy vehicle electric drive assembly energy consumption optimization system, used to implement the new energy vehicle electric drive assembly energy consumption optimization method as described in any one of claims 1 to 5, characterized in that, include: Information acquisition module: Acquires road and traffic information for the vehicle's preset route ahead; Torque planning module: Based on the road and traffic information, predicts the vehicle's power demand during future driving cycles and generates a total demand torque command sequence accordingly; Parameter identification module: During the current operation of the vehicle, the permanent magnet synchronous motor is identified online in real time to obtain the real-time electromagnetic parameters of the permanent magnet synchronous motor; Curve correction module: Dynamically corrects the reference efficiency mapping curve based on the real-time electromagnetic parameters; Command synthesis module: Based on the corrected benchmark efficiency mapping curve, the total demand torque command sequence is decomposed and synthesized into real-time control commands for each permanent magnet synchronous motor. Drive control module: Executes the real-time control commands to drive the permanent magnet synchronous motor.

Images