A data-driven based vehicle trip prediction and foresighted energy management method

Abstract

The present application belongs to the technical field of electric vehicle thermal management control, and in particular to a vehicle trip prediction and predictive energy management method based on data driving. The method comprises the following steps: S1: data acquisition and feature construction; S2: time sequence working condition prediction; S3: energy demand assessment; S4: predictive energy management decision; and S5: control instruction feedforward execution. The present application reduces the SOC target value of the battery in advance through deceleration feedback predictability control before entering a long downhill or a red-green light, and releases the receiving space in advance; the battery is cooled in advance before a high-power working condition arrives, the thermal inertia of the battery is utilized, and the battery temperature is prevented from exceeding the optimal interval during operation, thereby effectively avoiding high-temperature power limitation and prolonging the battery life. The present application dynamically feeds forward to limit the air conditioner and heating accessory power in combination with the remaining time of the trip and the environmental temperature, thereby avoiding excessive consumption of vehicle energy and realizing the collaborative energy saving of the driving system and the high-voltage accessories.

Description

Technical Field

[0001] This invention relates to the field of electric vehicle thermal management control technology, specifically to a data-driven method for vehicle trip prediction and predictive energy management. Background Technology

[0002] With the rapid development of electric vehicles, improving overall vehicle energy efficiency, extending driving range, and ensuring the safety and lifespan of energy storage systems have become critical issues that urgently need to be addressed in the field of vehicle engineering. Vehicle energy management strategies, as the core of regulating the power distribution among various vehicle components, have a decisive impact on the overall vehicle economy and performance.

[0003] Traditional vehicle energy management strategies rely primarily on reactive adjustments based on the vehicle's current real-time state (such as current speed and accelerator pedal position). However, this reactive control strategy cannot anticipate future driving conditions. For example, when a vehicle is about to encounter a long downhill section, traditional strategies cannot pre-emptively deplete the battery to create regenerative braking capacity, resulting in unrecoverable braking energy during the downhill as the battery is fully charged. Similarly, before facing high-power demand conditions such as continuous uphill climbing, the system cannot proactively cool the battery in advance, causing the battery to overheat rapidly during the climb, triggering power limiting protection and severely impacting vehicle performance and battery life.

[0004] Although there are some existing methods for trip prediction based on navigation maps, these methods often lack deep integration of real-time sensor data, drivers' personalized driving habits, and high-precision three-dimensional slope information, resulting in low accuracy of condition prediction and an inability to achieve refined predictive control.

[0005] In view of this, how to deeply integrate multi-source driving environment data with vehicle dynamics state, construct a high-precision future time-series operating condition prediction model, and on this basis realize predictive collaborative control that integrates drive power distribution, active thermal management of energy storage system and accessory power consumption limitation is a major technical bottleneck currently facing the field of vehicle energy management. Summary of the Invention

[0006] The purpose of this section is to outline some aspects of the embodiments of the present invention and to briefly describe some preferred embodiments. Simplifications or omissions may be made in this section, as well as in the abstract and title of this application, to avoid obscuring the purpose of these documents; however, such simplifications or omissions should not be construed as limiting the scope of the invention.

[0007] To address the aforementioned technical problems, according to one aspect of the present invention, the present invention provides the following technical solution:

[0008] A data-driven method for vehicle trip prediction and predictive energy management includes the following steps:

[0009] S1: Data Acquisition and Feature Construction: Acquire multi-source real-time data through vehicle sensors and external information sources, including vehicle status data, driver operation data, and travel environment data; extract features from the multi-source real-time data to construct an input feature vector;

[0010] S2: Time-series driving condition prediction: The input feature vector is input into a pre-trained trip prediction model to predict the driving condition sequence of the vehicle in the future time domain. The driving condition sequence includes at least the future vehicle speed time-series curve and the future road slope sequence.

[0011] S3: Energy demand assessment: Based on the driving condition sequence, combined with the vehicle dynamics model and power loss model, calculate the predicted energy demand distribution and power demand sequence of the vehicle in the future preset time period;

[0012] S4: Predictive Energy Management Decision: Taking minimizing the overall energy consumption of the vehicle and protecting battery life as the multi-objective optimization function, combined with the current state of charge of the power battery, battery temperature and the power demand sequence in the future time domain interval, the vehicle energy distribution parameters are dynamically adjusted in real time through a rolling optimization algorithm to generate a target torque correction coefficient for adjusting the drive torque and an accessory power consumption limit instruction for limiting accessory energy consumption.

[0013] S5: Control command feedforward execution: The target torque correction coefficient and accessory power consumption limit command are sent to the vehicle controller, which then performs feedforward coordinated control on the power distribution of the motor drive system and the on-board high-voltage accessories.

[0014] In a preferred embodiment of the data-driven vehicle trip prediction and predictive energy management method of the present invention, the on-board sensor in S1 includes:

[0015] Wheel speed sensors, inertial measurement units, and battery management system sampling chips are used to sense vehicle dynamics and energy status.

[0016] Accelerator pedal position sensor, brake pedal pressure sensor and steering wheel angle sensor are used to sense the driver's operating intentions.

[0017] And onboard cameras and radar used to sense the microscopic physical environment around the vehicle.

[0018] As a preferred embodiment of the data-driven vehicle trip prediction and predictive energy management method described in this invention, the external information sources in S1 include: data interaction with a cloud server through a built-in vehicle wireless communication terminal to retrieve high-precision elevation map data of the current planned route, real-time traffic flow congestion index data, and meteorological environment data of the current area.

[0019] As a preferred embodiment of the data-driven vehicle trip prediction and predictive energy management method described in this invention, the trip prediction model in S2 is a deep neural network model. The deep neural network model establishes a nonlinear mapping from multidimensional input features to future speed time series by learning historical driving trajectories and corresponding environmental features at the time.

[0020] As a preferred embodiment of the data-driven vehicle trip prediction and predictive energy management method described in this invention, in step S3, the calculation parameters input to the vehicle dynamics model include at least the real-time slope value calculated by the inertial measurement unit, or the future slope value obtained by matching high-precision elevation map data.

[0021] As a preferred embodiment of the data-driven vehicle trip prediction and predictive energy management method of the present invention, the predictive energy management decision in step S4 includes the following predictive control logic:

[0022] Deceleration feedback predictive control: When the predicted operating conditions indicate that there will be a continuous downhill or traffic light deceleration condition in the future time domain, the target value of the current state of charge of the power battery is actively reduced before the vehicle reaches the said operating conditions, and the battery capacity is released in advance to maximize the recovery of subsequent braking energy.

[0023] Thermal management predictive control: When the predicted operating conditions indicate that there will be continuous ramp-up or high-speed, high-power demand conditions in the future time domain, the battery cooling system will be turned on in advance to reduce the battery operating temperature to a preset optimal operating temperature range that matches the current battery type before entering the high-power demand conditions. The optimal operating temperature range is set differently according to the chemical system type of the power battery.

[0024] As a preferred embodiment of the data-driven vehicle trip prediction and predictive energy management method described in this invention, the on-board high-voltage accessories in S5 include an air conditioning system and a PTC heater; the feedforward cooperative control refers to dynamically adjusting the rated operating power limit of the air conditioning system and the PTC heater based on the remaining trip time and ambient temperature in the predicted operating conditions, so as to reduce non-driving energy consumption; the dynamic adjustment is constrained by the thermal comfort index of the passenger compartment.

[0025] Compared with existing technologies, the beneficial effects of this invention are as follows: This invention integrates microscopic dynamic data and driver control data collected by onboard sensors with macroscopic high-precision map data and traffic flow data obtained from external information sources, enabling future speed and gradient predictions to possess not only spatial accuracy but also temporal dynamic adaptability. Through deceleration feedback predictive control, this invention proactively reduces the battery's SOC target value before entering long downhill slopes or traffic lights, releasing the receiving space in advance and avoiding the pain point of traditional vehicles being unable to recover braking energy due to excessively high SOC. This invention pre-cools the battery before high-power operating conditions arrive, utilizing the battery's thermal inertia to prevent the battery temperature from exceeding the optimal range during operation, effectively avoiding high-temperature power limiting and extending battery life. This invention combines remaining travel time and ambient temperature to dynamically feedforward and limit the power of air conditioning and heating accessories, avoiding excessive consumption of onboard energy and achieving coordinated energy saving of the drive system and high-voltage accessories. Attached Figure Description

[0026] To more clearly illustrate the technical solutions of the embodiments of the present invention, the present invention will be described in detail below with reference to the accompanying drawings and detailed embodiments. 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:

[0027] Figure 1 This is a schematic diagram of the overall process in an embodiment of a data-driven vehicle trip prediction and predictive energy management method of the present invention;

[0028] Figure 2 This is a schematic diagram of a multi-source data fusion and trip prediction model in an embodiment of a data-driven vehicle trip prediction and predictive energy management method of the present invention;

[0029] Figure 3 This is an example of the predictive energy management and feedforward control process in an embodiment of a data-driven vehicle trip prediction and predictive energy management method of the present invention. Detailed Implementation

[0030] To make the above-mentioned objects, features and advantages of the present invention more apparent and understandable, the specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings.

[0031] Secondly, the present invention is described in detail with reference to the schematic diagrams. When detailing the embodiments of the present invention, for ease of explanation, the cross-sectional views illustrating the device structure may be partially enlarged, not according to the usual scale. Furthermore, the schematic diagrams are merely examples and should not limit the scope of protection of the present invention. In addition, actual fabrication should include three-dimensional spatial dimensions of length, width, and depth.

[0032] To make the objectives, technical solutions, and advantages of the present invention clearer, the embodiments of the present invention will be described in further detail below with reference to the accompanying drawings.

[0033] A data-driven method for vehicle trip prediction and predictive energy management includes the following steps:

[0034] Step S1: Data Acquisition and Feature Construction: Acquire multi-source real-time data through vehicle sensors and external information sources, including vehicle status data, driver operation data, and travel environment data; extract features from the multi-source real-time data to construct an input feature vector;

[0035] The onboard sensors include: wheel speed sensors, inertial measurement units (IMUs), and battery management system (BMS) sampling chips for sensing vehicle dynamics and energy state; accelerator pedal position sensors, brake pedal pressure sensors, and steering wheel angle sensors for sensing driver intentions; and onboard cameras and radar for sensing the microscopic physical environment around the vehicle.

[0036] External information sources include: data interaction with the cloud server through the built-in vehicle wireless communication terminal (T-Box) to retrieve high-precision elevation map data of the current planned route, real-time traffic flow congestion index data, and meteorological environmental data of the current area;

[0037] Step S2: Time-series driving condition prediction: Input the input feature vector into the pre-trained trip prediction model to predict the driving condition sequence of the vehicle in the future time domain. The driving condition sequence includes at least the future vehicle speed time-series curve and the future road slope sequence.

[0038] The trip prediction model is a deep neural network model. The model learns the historical driving trajectory and the environmental features at the corresponding time to establish a nonlinear mapping from multi-dimensional input features to future speed time series.

[0039] Step S3: Energy demand assessment: Based on the driving condition sequence, combined with the vehicle dynamics model and power loss model, calculate the predicted energy demand distribution and power demand sequence of the vehicle in the future preset time period;

[0040] The vehicle dynamics model, whose input calculation parameters include at least the real-time slope value calculated by the inertial measurement unit (IMU) or the future slope value obtained by matching high-precision elevation map data;

[0041] Step S4: Predictive Energy Management Decision: Taking minimizing the overall energy consumption of the vehicle and protecting battery life as the multi-objective optimization function, the vehicle energy distribution parameters are dynamically adjusted in real time through a rolling optimization algorithm, combined with the current state of charge (SOC) of the power battery, battery temperature and the power demand sequence in the future time domain, to generate a target torque correction coefficient for adjusting the drive torque and an accessory power consumption limit instruction for limiting accessory energy consumption.

[0042] Predictive energy management decisions include the following predictive control logic:

[0043] Deceleration feedback predictive control: When the predicted operating conditions indicate that there will be a continuous downhill or traffic light deceleration condition in the future time domain, the target value of the current state of charge of the power battery is actively reduced before the vehicle reaches the operating condition, and the battery capacity is released in advance to maximize the recovery of subsequent braking energy.

[0044] Thermal management predictive control: When the predicted operating conditions indicate that there will be continuous ramp-up or high-speed, high-power demand conditions in the future time domain, the battery cooling system will be turned on in advance to reduce the battery operating temperature to the preset optimal operating temperature range that matches the current battery type before entering the high-power demand conditions. The optimal operating temperature range is set differently according to the chemical system type of the power battery.

[0045] Step S5: Control command feedforward execution: The target torque correction coefficient and accessory power consumption limit command are sent to the vehicle controller (VCU), and the vehicle controller performs feedforward coordinated control of the power distribution between the motor drive system and the on-board high-voltage accessories;

[0046] The on-board high-voltage accessories include the air conditioning system and the PTC heater; feedforward cooperative control refers to: dynamically adjusting the upper limit of the rated operating power of the air conditioning system and the PTC heater based on the predicted remaining travel time and ambient temperature in the operating conditions, so as to reduce non-driving energy consumption; the dynamic adjustment is constrained by the thermal comfort index of the passenger compartment.

[0047] Example:

[0048] like Figure 1 As shown, the overall flowchart of a data-driven vehicle trip prediction and predictive energy management method provided by the present invention includes:

[0049] S1: Acquire real-time data from multiple sources through onboard sensors and external information sources.

[0050] 1. Vehicle-mounted sensor data acquisition:

[0051] The vehicle's real-time speed is obtained through wheel speed sensors. ;

[0052] Obtain the vehicle's real-time longitudinal acceleration Real-time longitudinal acceleration The acquisition method depends on the real-time speed of the data. The result is obtained by taking the first derivative with respect to time t, i.e.:

[0053]

[0054] The inertial measurement unit acquires triaxial acceleration and angular velocity to calculate the vehicle's current real-time road slope. ;

[0055] The battery management system uses a chip to obtain the current state of charge of the power battery. Battery temperature , Single cell maximum / minimum voltage and battery state of health (SOH);

[0056] The driver's input characteristics are obtained by using the accelerator pedal position sensor, brake pedal pressure sensor and steering wheel angle sensor to identify the driver's driving style (such as aggressive, standard and economical).

[0057] By using vehicle-mounted cameras and radar to acquire information such as distance to other vehicles, relative speed to the vehicle in front, and obstacles ahead, the vehicle can perceive the microscopic physical environment.

[0058] 2. External information source data collection:

[0059] The vehicle interacts with the cloud server bidirectionally via a built-in in-vehicle wireless communication terminal (T-Box).

[0060] Based on the current planned route of the vehicle navigation system, high-precision map data within a 2km radius ahead is retrieved in advance (to extract future road slope sequences). Real-time traffic congestion index data and current regional meteorological data. .

[0061] 3. Feature Reconstruction:

[0062] The collected multi-source real-time data is time-aligned and normalized. An input feature vector is then constructed.

[0063]

[0064] like Figure 2 The diagram illustrates a multi-source data fusion and trip prediction model using a data-driven vehicle trip prediction and predictive energy management approach. The specific steps include:

[0065] Input feature vector The data is input into a pre-trained trip prediction model. The trip prediction model employs a deep neural network model that combines a long short-term memory network with an attention mechanism.

[0066] Temporal mapping features: The model uses historical 30-second sliding window data as input sequence. By learning the historical driving trajectory and the environmental features at the corresponding time, it establishes a nonlinear mapping from multi-dimensional input features to future speed time series.

[0067] Model Output: Outputs the driving condition sequence within the future time domain. The driving condition sequence must include at least the future vehicle speed time-series curve:

[0068]

[0069] The future road gradient sequence is as follows:

[0070]

[0071] like Figure 3 The diagram shows the flowchart of predictive energy management and feedforward control using a data-driven vehicle trip prediction and predictive energy management method. Specifically, it includes the following steps:

[0072] Based on the predicted driving condition sequence, combined with the vehicle dynamics model and power loss model, the predicted energy demand distribution and power demand sequence of the vehicle in the future preset time period are calculated.

[0073] (1) Vehicle power model:

[0074] Vehicle demand drivers The calculation formula is:

[0075]

[0076] in, For the overall vehicle weight; The acceleration due to gravity is taken as 9.8 m / s². 2 ; This is the rolling resistance coefficient, with a value ranging from 0.005 to 0.04. The future road surface slope predicted in step S2 has a value range of -18° to 18°. This is the drag coefficient, with a value ranging from 0.2 to 0.8. The windward area is measured in values ​​ranging from 1.5 to 11 m². 2 ; This refers to air density, with a value ranging from 0.8 to 1.4 kg / m³. 3 ; To predict vehicle speed, This is the rotational mass conversion factor, with a value range of 1.01-1.3.

[0077] (2) Power loss model:

[0078] Vehicle motor power requirements for:

[0079] when (Driving condition)

[0080]

[0081] when (During braking conditions):

[0082]

[0083] in, For the overall efficiency of motor drive and transmission systems. This represents the overall efficiency of regenerative braking. These two efficiency values ​​are calculated by looking up a pre-stored motor efficiency map.

[0084] Therefore, the power demand sequence within the future time domain can be calculated:

[0085]

[0086] To extract the statistical characteristics of future power demand, a power demand probability density distribution model was constructed. The specific steps are as follows:

[0087] R1 regenerative braking zone: ;

[0088] R2 low power region: (10%-20% of the rated power of the drive motor);

[0089] R3 Economic Operation Zone: (60%-80% of the rated power of the drive motor)

[0090] R4 high load area:

[0091] For N predicted points within the future time domain T, the probability of falling into the j-th interval. The calculation can be expressed as:

[0092]

[0093] in, The predicted energy demand distribution for the whole vehicle, which is an indicator function (1 when the condition is met, 0 otherwise), is obtained by summing the demand power series over the future time domain interval. Its calculation formula can be expressed as:

[0094]

[0095] in, For the preset adoption period, This represents the cumulative energy requirement from the current moment until the kth sampling point in the future.

[0096] Based on the rolling optimization algorithm, the vehicle energy distribution parameters are dynamically adjusted in real time. The optimization function is to minimize the overall energy consumption of the vehicle and protect the battery life. The optimal power flow distribution strategy is solved by combining the current state of charge (SOC) of the power battery, the battery temperature and the power demand sequence in the future time domain. The target torque correction coefficient and accessory power consumption limit instructions are generated.

[0097] Within this optimized framework, the following core predictive control logic is designed in this embodiment:

[0098] (1) Deceleration feedback predictive control:

[0099] The predicted working conditions indicate a continuous downhill slope (gradient) in the future time domain. When the vehicle reaches a speed of -4° or deceleration at traffic lights, if the current SOC of the power battery is at a high level (SOC≥80%), the vehicle controller will proactively lower the target value of the current power battery's state of charge before the vehicle reaches the operating condition. This is achieved by increasing the motor drive output and sharing the consumption of high-voltage accessories, thus releasing the receiving capacity of the energy storage system in advance. When the vehicle actually reaches a downhill or deceleration condition, the freed-up space can maximize the recovery of braking energy from the downhill slope.

[0100] (2) Predictive control of thermal management:

[0101] When the predicted working conditions indicate that there will be a continuous uphill slope (gradient) in the future time domain. When the temperature exceeds 5°C (for up to 1 km) or during high-speed, high-power demand conditions, the system will activate the battery cooling system in advance to reduce the battery operating temperature to the preset optimal operating temperature range before entering high-power demand conditions.

[0102] Specific settings for the optimal temperature range:

[0103] The power battery is a ternary lithium battery, with the optimal operating temperature range preset to 25-35℃;

[0104] If the power battery is a lithium iron phosphate battery, the optimal operating temperature range is preset to 20-40℃.

[0105] If the power battery is a sodium-ion battery, the optimal operating temperature range is preset to 15-35℃.

[0106] By pre-cooling, the battery temperature remains within the efficient operating range during the high-temperature rise phase of the vehicle's ascent, avoiding power limiting protection caused by the battery triggering the high-temperature threshold, ensuring power continuity and extending battery life.

[0107] This embodiment designs the following control parameters for generation:

[0108] Target torque correction factor The value ranges from 0.8 to 1, and is used to actively limit the torque demand of the drive motor when future energy demand is too high or the battery level is low, thereby extending the driving range. The target torque correction changes smoothly within the filter time constant to ensure the smoothness of the vehicle's longitudinal dynamic response.

[0109]

[0110] in, This represents the difference between the reference SOC and the actual SOC at the current moment; The probability of occurrence of the high-load region R4 in the power probability distribution vector; To correct the sensitivity factor, the value range is 0.1-0.3, and 0.2 is selected in this study. To correct the weighting function, it monotonically increases with the increase of the power deviation, and its value range is [0,1].

[0111] Based on the remaining travel time and the remaining available battery energy, dynamically plan the power limit of high-voltage accessories (air conditioner, PTC):

[0112] in, This represents the remaining usable energy at the current battery level. The estimated remaining time for the trip; The preset drive power allocation ratio is 0.8; This is the temperature environment regulation coefficient, used to ensure basic comfort in the passenger cabin under extreme temperatures. (When the ambient temperature is greater than 38℃ or less than -15℃, the value is 1.2; when the ambient temperature is between 15-25℃, the value is 0.5; and in other temperature ranges, the value is 1.) The core predictive control logic is qualitative intervention, and the solution of the control parameters is quantitative correction. Together, they constitute the predictive energy management decision.

[0113] The vehicle control unit (VCU) receives the generated... and Real-time instruction delivery to the execution layer hardware:

[0114] (1) Motor drive feedforward execution: The vehicle controller obtains the driver's current pedal torque requirement. ; Target torque reshaping is performed using the generated correction coefficients: ; the revised The signal is sent to the motor controller (MCU) to achieve early avoidance of future high-energy-consumption operating conditions, while ensuring that the longitudinal dynamic response deviation of the vehicle is within the preset tolerance range.

[0115] (2) High-voltage accessory power consumption limit execution: The vehicle controller will The command is sent to the air conditioner controller or PTC control module via the CAN bus; after receiving the limiting command, the air conditioner controller reduces the compressor speed or adjusts the fan power to suppress the total power consumption. Within the specified range; in low-temperature environments, the power of the heating rod is reduced by adjusting the duty cycle to ensure that the energy consumption of the accessories does not exceed the planned upper limit.

[0116] (3) Closed-loop feedback and dynamic update: The vehicle controller monitors the rate of change of SOC after execution in real time. If the SOC deviation returns to the normal range, it will be gradually released in the next calculation cycle. and The limitations were removed, restoring normal power output and comfort levels.

[0117] Although the present invention has been described above with reference to embodiments, various modifications can be made and components can be replaced with equivalents without departing from the scope of the invention. In particular, as long as there is no structural conflict, the features in the disclosed embodiments can be combined with each other in any manner. The lack of an exhaustive description of these combinations in this specification is merely for the sake of brevity and resource conservation. Therefore, the present invention is not limited to the specific embodiments disclosed herein, but includes all technical solutions falling within the scope of the claims.

Claims

1. A data-driven method for vehicle trip prediction and predictive energy management, characterized in that, Includes the following steps: S1: Data Acquisition and Feature Construction: Acquire multi-source real-time data through vehicle sensors and external information sources, including vehicle status data, driver operation data, and travel environment data; extract features from the multi-source real-time data to construct an input feature vector; S2: Time-series driving condition prediction: The input feature vector is input into a pre-trained trip prediction model to predict the driving condition sequence of the vehicle in the future time domain. The driving condition sequence includes at least the future vehicle speed time-series curve and the future road slope sequence. S3: Energy demand assessment: Based on the driving condition sequence, combined with the vehicle dynamics model and power loss model, calculate the predicted energy demand distribution and power demand sequence of the vehicle in the future preset time period; S4: Predictive Energy Management Decision: Taking minimizing the overall energy consumption of the vehicle and protecting battery life as the multi-objective optimization function, combined with the current state of charge of the power battery, battery temperature and the power demand sequence in the future time domain interval, the vehicle energy distribution parameters are dynamically adjusted in real time through a rolling optimization algorithm to generate a target torque correction coefficient for adjusting the drive torque and an accessory power consumption limit instruction for limiting accessory energy consumption. S5: Control command feedforward execution: The target torque correction coefficient and accessory power consumption limit command are sent to the vehicle controller, which then performs feedforward coordinated control on the power distribution of the motor drive system and the on-board high-voltage accessories.

2. The data-driven vehicle trip prediction and predictive energy management method according to claim 1, characterized in that, The on-board sensors in S1 include: Wheel speed sensors, inertial measurement units, and battery management system sampling chips are used to sense vehicle dynamics and energy status. Accelerator pedal position sensor, brake pedal pressure sensor and steering wheel angle sensor are used to sense the driver's operating intentions. And onboard cameras and radar used to sense the microscopic physical environment around the vehicle.

3. The data-driven vehicle trip prediction and predictive energy management method according to claim 1, characterized in that, The external information sources in S1 include: data interaction with the cloud server through the built-in vehicle wireless communication terminal to retrieve high-precision elevation map data of the current planned route, real-time traffic flow congestion index data, and meteorological environment data of the current area.

4. The data-driven vehicle trip prediction and predictive energy management method according to claim 1, characterized in that, The trip prediction model in S2 is a deep neural network model. The deep neural network model learns the historical driving trajectory and the environmental features at the corresponding time to establish a nonlinear mapping from multi-dimensional input features to future speed time series.

5. The data-driven vehicle trip prediction and predictive energy management method according to claim 1, characterized in that, In S3, the calculation parameters input to the vehicle dynamics model include at least the real-time slope value calculated by the inertial measurement unit, or the future slope value obtained by matching high-precision elevation map data.

6. The data-driven vehicle trip prediction and predictive energy management method according to claim 1, characterized in that, The predictive energy management decision in step S4 includes the following predictive control logic: Deceleration feedback predictive control: When the predicted operating conditions indicate that there will be a continuous downhill or traffic light deceleration condition in the future time domain, the target value of the current state of charge of the power battery is actively reduced before the vehicle reaches the said operating conditions, and the battery capacity is released in advance to maximize the recovery of subsequent braking energy. Thermal management predictive control: When the predicted operating conditions indicate that there will be continuous ramp-up or high-speed, high-power demand conditions in the future time domain, the battery cooling system will be turned on in advance to reduce the battery operating temperature to a preset optimal operating temperature range that matches the current battery type before entering the high-power demand conditions. The optimal operating temperature range is set differently according to the chemical system type of the power battery.

7. The data-driven vehicle trip prediction and predictive energy management method according to claim 1, characterized in that, The on-board high-voltage accessories in S5 include an air conditioning system and a PTC heater; the feedforward cooperative control refers to dynamically adjusting the rated operating power limit of the air conditioning system and the PTC heater based on the predicted remaining travel time and ambient temperature in the operating condition, so as to reduce non-driving energy consumption; the dynamic adjustment is constrained by the thermal comfort index of the passenger compartment.

Images