Vehicle thermal management method and device, electronic equipment and storage medium

Abstract

The invention relates to the technical field of vehicles, and discloses a vehicle thermal management method and device, electronic equipment and a storage medium. The method comprises the steps of at least obtaining state information of a vehicle; at least establishing a battery domain thermal load prediction model, a motor domain thermal load prediction model, an electric control domain thermal load prediction model, a passenger cabin domain thermal load prediction model, a navigation working condition identification model, a power demand prediction model and an environment change prediction model of the vehicle on the basis of a preset algorithm architecture and each piece of state information; a vehicle multi-objective optimization function and constraint conditions thereof are constructed according to the state information; and according to each model, the whole vehicle multi-objective optimization function and the constraint condition thereof, executing whole vehicle thermal management cooperative operation. According to the application, cooperative control of four thermal management domains of a battery, a motor, electric control and a passenger cabin in the new energy commercial vehicle can be at least realized, so that multi-objective optimization of the energy efficiency of the whole vehicle, the service life of the system and the response speed is realized on the premise of ensuring the temperature control precision of each domain, and the driving experience of a user is favorably ensured.

Description

Technical Field

[0001] This invention relates to the field of vehicle technology, and in particular to a vehicle thermal management method, device, electronic device, and storage medium. Background Technology

[0002] With the rapid development of new energy commercial vehicles, the thermal management system, as a key system affecting vehicle energy efficiency, battery life, and driving comfort, is becoming increasingly important. Traditional thermal management systems employ a separate control approach, with battery thermal management, motor thermal management, electronic control thermal management, and cabin air conditioning operating independently, lacking coordinated optimization and exhibiting at least the following problems:

[0003] 1. Low energy efficiency: Each subsystem operates independently, making it impossible to effectively transfer and utilize heat. For example, the waste heat generated by the motor cannot be used for cabin heating or battery preheating.

[0004] 2. Control strategy conflict: The control objectives of different subsystems may conflict with each other. For example, when the battery cooling demand and the cabin cooling demand exist at the same time, the compressor load is too high.

[0005] 3. System response lag: The lack of predictive control means that the system can only passively respond to temperature changes, leading to increased energy consumption and decreased comfort. Summary of the Invention

[0006] The purpose of this invention is to provide a vehicle thermal management method, device, electronic device and storage medium, which can at least realize the coordinated control of multiple thermal management domains in new energy commercial vehicles, so as to achieve multi-objective optimization of vehicle energy efficiency, system life and response speed while ensuring the temperature control accuracy of each domain, thereby helping to protect the user's driving experience.

[0007] To address the aforementioned technical problems, in a first aspect, the present invention provides a vehicle thermal management method, which is at least applicable to new energy commercial vehicles;

[0008] The vehicle thermal management method includes at least the following:

[0009] At least the vehicle's battery domain status information, motor domain status information, electronic control domain status information, passenger cabin domain status information, and overall vehicle operating status information should be obtained;

[0010] Based on the preset algorithm architecture and each state information, at least the following models are established for predicting vehicle thermal load: battery domain thermal load, motor domain thermal load, electronic control domain thermal load, cabin domain thermal load, navigation condition identification, power demand prediction, and environmental change prediction. A multi-objective optimization function for the whole vehicle and its constraints are constructed based on the state information.

[0011] The vehicle thermal management collaborative operation is performed based on each of the aforementioned heat load prediction model, navigation condition identification model, power demand prediction model, environmental change prediction model, vehicle multi-objective optimization function, and their constraints.

[0012] Based on the same concept, in a second aspect, the present invention also provides a vehicle thermal management device for performing the vehicle thermal management method described in any one of the first aspects;

[0013] The vehicle thermal management device includes at least:

[0014] The information acquisition module is used to acquire at least the vehicle's battery domain status information, motor domain status information, electronic control domain status information, passenger cabin domain status information, and overall vehicle operating status information.

[0015] The model building module is used to build at least the following models based on the preset algorithm architecture and each state information: battery domain thermal load prediction model, motor domain thermal load prediction model, electronic control domain thermal load prediction model, cabin domain thermal load prediction model, navigation condition identification model, power demand prediction model and environmental change prediction model, and to construct the vehicle multi-objective optimization function and its constraints based on the state information.

[0016] The collaborative management module is used to perform collaborative operations for vehicle thermal management based on each of the aforementioned thermal load prediction models, navigation condition identification models, power demand prediction models, environmental change prediction models, vehicle multi-objective optimization functions, and their constraints.

[0017] Based on the same concept, in a third aspect, the present invention also provides an electronic device including a memory and a processor, the memory storing a computer program executable on the processor, the processor executing the program to implement the steps of the vehicle thermal management method described in any one of the first aspects.

[0018] Based on the same concept, in a fourth aspect, the present invention also provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the steps of the vehicle thermal management method described in any one of the first aspects.

[0019] The technical solution provided by this invention firstly acquires at least the battery domain state information, motor domain state information, electronic control domain state information, passenger cabin domain state information, and overall vehicle operating status information. Further, based on a preset algorithm architecture and each state information, at least the following models are established: battery domain thermal load prediction model, motor domain thermal load prediction model, electronic control domain thermal load prediction model, passenger cabin thermal load prediction model, navigation condition identification model, power demand prediction model, and environmental change prediction model. A multi-objective optimization function for the entire vehicle and its constraints are then constructed based on the state information. Finally, coordinated operation of the vehicle's thermal management is performed according to each thermal load prediction model, navigation condition identification model, power demand prediction model, environmental change prediction model, and the multi-objective optimization function for the entire vehicle and its constraints. Therefore, this invention can at least achieve coordinated control of the four major thermal management domains—battery, motor, electronic control, and passenger cabin—in new energy commercial vehicles. This ensures the accuracy of temperature control in each domain while achieving multi-objective optimization of vehicle energy efficiency, system lifespan, and response speed, thus improving the user's driving experience. Attached Figure Description

[0020] Figure 1 This is a flowchart of a vehicle thermal management method provided in an embodiment of the present invention;

[0021] Figure 2 This is a schematic diagram of the structure of a vehicle thermal management device provided in an embodiment of the present invention;

[0022] Figure 3 This is a schematic diagram of the structure of an electronic device provided in an embodiment of the present invention. Detailed Implementation

[0023] To make the objectives, technical solutions, and advantages of this application clearer, the application will be further described in detail below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments. Based on the embodiments in this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0024] The terminology used in the embodiments of this application is for the purpose of describing particular embodiments only and is not intended to limit the application. The singular forms “a,” “said,” and “the” used in the embodiments of this application and the appended claims are also intended to include the plural forms, and “multiple” generally includes at least two unless the context clearly indicates otherwise.

[0025] It should be understood that the term "and / or" used in this article is merely a description of the relationship between related objects, indicating that three relationships can exist. For example, A and / or B can represent: A existing alone, A and B existing simultaneously, and B existing alone. Additionally, the character " / " in this article generally indicates that the preceding and following related objects have an "or" relationship.

[0026] It should be understood that although the terms first, second, third, etc., may be used in the embodiments of this application, these descriptions should not be limited to these terms. These terms are only used to distinguish the descriptions. For example, first may also be referred to as second without departing from the scope of the embodiments of this application, and similarly, second may also be referred to as first.

[0027] Depending on the context, the words “if” or “suppose” as used here can be interpreted as “when” or “in response to determination” or “in response to detection.” Similarly, depending on the context, the phrases “if determination” or “if detection (of the stated condition or event)” can be interpreted as “when determination” or “in response to determination” or “when detection (of the stated condition or event)” or “in response to detection (of the stated condition or event).”

[0028] It should also be noted that the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that an article or device that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such an article or device. Without further limitation, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the article or device that includes said element.

[0029] It should be noted that any symbols and / or numbers present in the specification that are not marked in the accompanying drawings are not reference numerals.

[0030] Figure 1 This is a flowchart of a vehicle thermal management method provided by an embodiment of the present invention. This embodiment is applicable to at least any integrated thermal management scenario for new energy medium and heavy-duty commercial vehicles. The vehicle thermal management method can be, but is not limited to, executed by the vehicle thermal management device in this embodiment as the executing entity, which can be implemented in software and / or hardware. Figure 1 As shown, the vehicle thermal management method includes at least the following steps:

[0031] S1. Obtain at least the vehicle's battery domain status information, motor domain status information, electronic control domain status information, passenger cabin domain status information, and overall vehicle operating status information.

[0032] Among them, battery domain status information may include cell temperature distribution, SOC, charging and discharging power, battery coolant temperature, etc.; motor domain status information may include stator temperature, rotor temperature, motor coolant temperature, operating power, etc.; electronic control domain status information may include IGBT junction temperature, capacitor temperature, radiator temperature, etc.; cabin domain status information may include vehicle interior temperature, humidity, number of occupants, solar radiation intensity, etc.; and vehicle operating status information may include vehicle speed, acceleration, navigation route, ambient temperature, weather forecast, etc.

[0033] It is understandable that there are multiple ways to obtain the above status information. Specifically, coolant temperature, vehicle speed, etc., can be directly measured by sensors; cell temperature distribution, SOC, etc., can be calculated based on sensor measurement data; navigation routes, weather forecasts, etc., can be obtained through in-vehicle applications, which will not be elaborated further.

[0034] S2. Based on the preset algorithm architecture and each state information, at least the following models should be established for vehicle thermal load prediction: battery domain thermal load prediction model, motor domain thermal load prediction model, electronic control domain thermal load prediction model, cabin domain thermal load prediction model, navigation condition identification model, power demand prediction model, and environmental change prediction model. Based on the state information, construct the vehicle multi-objective optimization function and its constraints.

[0035] There are several options for the choice of the preset algorithm architecture. In one specific implementation, the preset algorithm architecture may optionally employ at least a recurrent neural network incorporating a long short-term memory layer.

[0036] Specifically, a recurrent neural network (RNN) combined with a long short-term memory (LSTM) layer is used to capture the temporal characteristics of heat load.

[0037] 1. Network structure:

[0038] Input layer → LSTM layer 1 → Dropout layer → LSTM layer 2 → Fully connected layer → Output layer.

[0039] (1) Input layer:

[0040] Input feature dimension: dim_input=15;

[0041] Feature list:

[0042] [v_vehicle,a_vehicle,SOC,T_bat,T_mot,T_inv,T_cab,T_amb,P_mot_current,I_bat,grade,traffic_flow,Q_solar,time_of_day,day_of_week];

[0043] Time step: timesteps=10 (using data from the past 5 minutes, one sample every 30 seconds).

[0044] (2) LSTM layer 1:

[0045] Number of hidden units: hidden_units_1=64;

[0046] Activation function: tanh;

[0047] Whether to return a sequence: return_sequences=True;

[0048] Function: Extract short-term time series features.

[0049] (3) Dropout layer:

[0050] Dropout rate: dropout_rate=0.2;

[0051] Function: Prevents overfitting.

[0052] (4) LSTM layer 2:

[0053] Number of hidden units: hidden_units_2=32;

[0054] Activation function: tanh;

[0055] Whether to return a sequence: return_sequences=False;

[0056] Function: Extract long-term dependencies.

[0057] (5) Fully connected layer:

[0058] Number of neurons: dense_units=64;

[0059] Activation function: ReLU;

[0060] Function: Nonlinear mapping.

[0061] (6) Output layer:

[0062] Number of neurons: dim_output=4;

[0063] Output content: [Q_bat_pred,Q_mot_pred,Q_inv_pred,Q_cab_pred];

[0064] Activation function: Linear (regression task).

[0065] 2. Training Data Preparation and Preprocessing

[0066] (1) Data collection:

[0067] from vehicle historical operation database:

[0068] Fields to be collected:

[0069] timestamp,v,a,SOC,T_bat,T_mot,T_inv,T_cab,T_amb,P_mot,I_bat,grade,Q_solar,Q_bat_actual,Q_mot_actual,...;

[0070] Acquisition cycle: 10Hz (one record every 0.1s);

[0071] Data volume: At least 10,000 km of operating mileage data.

[0072] (2) Data preprocessing:

[0073] Step 1: Downsampling

[0074] Downsampling from 10Hz to 0.033Hz (one line every 30s);

[0075] sample_data=raw_data[::300] # Take 1 data out of every 300 data entries.

[0076] Step 2: Data Normalization

[0077] Normalize to the [0,1] interval using Min-Max:

[0078] x_norm=(x-x_min) / (x_max-x_min).

[0079] Normalization range for each feature:

[0080] v_vehicle:0km / h-120km / h→[0,1];

[0081] SOC: 0-100% → [0,1];

[0082] T_bat:-20℃-60℃→[0,1];

[0083] T_amb: -40℃~60℃→[0,1];

[0084] P_mot: 0kW-300kW→[0,1];

[0085] ...

[0086] Step 3: Construct time series samples

[0087] foriinrange(len(sample_data)-timesteps-prediction_horizon):

[0088] X[i] = sample_data[i:i+timesteps] # Input: the past 10 time steps;

[0089] y[i]=sample_data[i+timesteps+prediction_horizon] # Output: 1 time step in the future;

[0090] prediction_horizon=10 (predict the next 5 minutes, 10 time steps × 30s).

[0091] Step 4: Dataset Partitioning

[0092] Training set: 70%;

[0093] Validation set: 15%;

[0094] Test set: 15%.

[0095] Organize data chronologically to prevent leaks:

[0096] train_data=data[:int(0.7*len(data))]

[0097] val_data=data[int(0.7*len(data)):int(0.85*len(data))]

[0098] test_data=data[int(0.85*len(data)):].

[0099] 3. Model Training Strategy

[0100] Offline training phase (can be completed in the development environment):

[0101] (1) Training parameters:

[0102] batch_size=64 (batch size);

[0103] epochs=100 (number of training epochs);

[0104] learning_rate=0.001 (initial learning rate);

[0105] optimizer = Adam (Adaptive learning rate optimizer);

[0106] loss_function = MSE (mean squared error).

[0107] (2) Learning rate decay strategy:

[0108] The learning rate decays by 50% every 20 epochs.

[0109] lr_schedule=lr_initial×0.5^(epoch / / 20).

[0110] (3) Early shutdown mechanism:

[0111] ifval_loss has not improved for 10 consecutive epochs:

[0112] Stop training and save the best model.

[0113] (4) Training process:

[0114] forepochin1to100:

[0115] forbatchintrain_data:

[0116] #Forward propagation

[0117] y_pred = model.forward(batch.X)

[0118] loss = MSE(y_pred, batch.y)

[0119] #Reverse propagation

[0120] gradients=compute_gradients(loss)

[0121] optimizer.update(model.parameters,gradients)

[0122] #Validation Set Evaluation

[0123] val_loss=evaluate(model,val_data)

[0124] ifval_loss <best_val_loss:

[0125] save_model(model,"best_model.h5").

[0126] (5) Model performance metrics (test set):

[0127] MAE (Mean Absolute Error): <0.5kW;

[0128] RMSE (Root Mean Square Error): <0.8kW;

[0129] R 2 (Determination coefficient): >0.92.

[0130] 4. Online learning update mechanism

[0131] After the model is deployed to the vehicle, it is continuously optimized through online learning:

[0132] (1) Triggering conditions for online learning:

[0133] ①Time-triggered: Updated every 100km driven or every 24 hours;

[0134] ②Performance trigger: Update immediately when the prediction error > threshold;

[0135] if |Q_actual - Q_pred| > 2.0kW:

[0136] trigger_online_learning().

[0137] (2) Online learning process:

[0138] Step 1: Data Collection

[0139] buffer=[] # Circular buffer, capacity 1000 entries

[0140] While the vehicle is running:

[0141] if sampling time:

[0142] record={features,actual_load}

[0143] buffer.append(record)

[0144] iflen(buffer)>1000:

[0145] buffer.pop(0) # Remove the oldest data.

[0146] Step 2: Incremental Training

[0147] if the trigger condition is met:

[0148] # Retrieve the latest 500 records from the buffer

[0149] recent_data=buffer[-500:]

[0150] #Small-batch fine-tuning (to avoid catastrophic oversight)

[0151] formini_epochin1to5:

[0152] formini_batchinrecent_data:

[0153] y_pred=model.forward(mini_batch.X)

[0154] loss = MSE(y_pred, mini_batch.y)

[0155] #Update with a smaller learning rate

[0156] lr_online=0.0001 # 10 times smaller than offline training

[0157] gradients=compute_gradients(loss)

[0158] optimizer.update(model.parameters,gradients,lr_online)

[0159] Step 3: Model Validation

[0160] #Verify on the most recent 100 data points

[0161] val_error=evaluate(updated_model,buffer[-100:])

[0162] ifval_error <current_model_error:

[0163] model=updated_model # Accept updates

[0164] save_model(model,"online_updated.h5")

[0165] else:

[0166] reject_update() # Reject the update and retain the original model.

[0167] Step 4: Model Version Management

[0168] model_versions={

[0169] "factory_default": Factory-pre-trained model.

[0170] "cloud_updated": The model is updated periodically in the cloud.

[0171] "local_adapted": Local online learning model

[0172] }

[0173] Priority: local_adapted>cloud_updated>factory_default

[0174] Iflocal_adapted performance degrades:

[0175] Revert to cloud_updated.

[0176] 5. Prediction accuracy verification and adaptive correction

[0177] (1) Real-time prediction error monitoring:

[0178] While the vehicle is running:

[0179] #Get the predicted value (the current value predicted 30 seconds ago).

[0180] Q_pred=prediction_buffer[current_time]

[0181] #Get the actual value (current measurement value)

[0182] Q_actual=measure_current_load()

[0183] #Calculation error

[0184] error=Q_actual-Q_pred

[0185] error_percentage=error / Q_actual×100%

[0186] #Record Error Statistics

[0187] error_history.append(error)

[0188] MAE_rolling=mean(abs(error_history[-100:]))# Rolling average absolute error

[0189] (2) Adaptive correction strategy:

[0190] If a systematic deviation is detected:

[0191] #Judgment criteria: 20 consecutive predicted values ​​are either too high or too low

[0192] ifall(error_history[-20:]>0.5):

[0193] bias=mean(error_history[-20:])

[0194] #Application Bias Correction

[0195] Q_pred_corrected = Q_pred + bias

[0196] #Simultaneously triggers online learning, fundamentally resolving biases.

[0197] trigger_online_learning()

[0198] (3) Selection of working condition specialization model:

[0199] Based on the current operating conditions, select the most suitable prediction model:

[0200] if operating condition == "high-speed cruising":

[0201] model=highway_model # Specifically designed for high-speed driving conditions

[0202] elif condition=="urban congestion":

[0203] model=urban_model # Specifically designed for urban working conditions

[0204] elif working condition == "Mountain road climbing":

[0205] model=mountain_model # Specifically designed for training on mountain roads

[0206] else:

[0207] model=general_model #General model.

[0208] In another specific implementation, the battery domain heat load prediction model may optionally be implemented in at least the following ways:

[0209] Q_bat=I 2 R_int+IVη_loss+k_conv(T_bat-T_amb);

[0210] In the above formula, Q_bat represents the battery heat load (unit: W, positive value indicates heat generation, negative value indicates heat absorption), I 2R_int represents the heat generated by the battery's internal resistance, I represents the battery's charging and discharging current (unit: A, range: -500A to +500A, positive value for discharging, negative value for charging), R_int represents the battery's equivalent internal resistance (unit: mΩ), IVη_loss represents the battery's charging and discharging loss, V represents the battery's charging and discharging voltage, η_loss represents the battery's loss coefficient (its value range: 0.02-0.05), k_conv(T_bat-T_amb) represents the battery's convective heat dissipation, k_conv represents the battery's convection coefficient (unit: W / ℃), T_bat represents the battery's average temperature (unit: ℃, operating range: -20℃ to 60℃), and T_amb represents the ambient temperature (unit: ℃, operating range: -40℃ to 60℃).

[0211] In yet another specific implementation, the motor domain heat load prediction model may optionally be implemented in at least the following ways:

[0212] Q_mot=P_mot(1-η_mot)+P_iron+P_mech;

[0213] In the above formula, Q_mot represents the motor thermal load (unit W, physical meaning is the continuous heat generated during motor operation), P_mot(1-η_mot) represents the motor copper loss and iron loss, P_mot represents the motor output power (unit kW, the value range of heavy truck is 0-250kW), η_mot represents the motor efficiency (its value range can be 0.92-0.96), P_iron represents the motor iron loss (which can be related to speed and magnetic flux density), and P_mech represents the motor mechanical loss (which can include bearing friction and wind resistance).

[0214] In yet another specific implementation, the electronically controlled domain heat load prediction model may optionally be implemented in at least the following ways:

[0215] Q_inv = P_sw + P_cond + P_cap;

[0216] In the above formula, Q_inv represents the electrical control heat load (unit: W, mainly from IGBT switching losses), P_sw represents the switching loss (P_sw = k_sw × P_inv × f_sw; where k_sw represents the switching loss coefficient; P_inv represents the inverter output power, unit: kW, ranging from 0kW to 300kW; f_sw represents the IGBT switching frequency, unit: kHz, typically 5kHz to 15kHz), and P_cond represents the conduction loss (P_cond = I 2 R_on (where R_on represents the on-resistance) and P_cap represents the capacitance loss (which can be related to the ripple current).

[0217] In yet another specific implementation, the cabin area heat load prediction model may optionally be implemented in at least the following ways:

[0218] Q_cab=Q_conv+Q_solar+Q_human+Q_equip;

[0219] In the above formula, Q_conv represents convective heat transfer (Q_conv=Ak_cab(T_amb-T_cab); where k_cab represents the vehicle body heat transfer coefficient, in W / (m³). 2 (℃), typical values ​​can be taken as 0.8-1.2; A represents the effective area of ​​the vehicle body, in m². 2 T_cab represents the actual cabin temperature in °C; T_amb represents the ambient temperature in °C; Q_solar represents solar radiation (i.e., solar heat radiation, which is related to the skylight area and glass transmittance, in W; it can reach 1000W-2000W on sunny days and 200W-500W on cloudy days); Q_human represents passenger heat dissipation (which can be estimated based on the number of passengers, approximately 100W per person); and Q_equip represents equipment heat dissipation (such as displays, charging equipment, etc.).

[0220] In yet another specific implementation, the multi-objective optimization function for the whole vehicle may optionally be implemented in at least the following ways:

[0221] J=w1E_total+w2ΔT_max+w3N_switch+w4T_response;

[0222] In the above formula, J represents the multi-objective optimization function of the whole vehicle, E_total represents the total energy consumption of the system, ΔT_max represents the maximum temperature difference in the battery domain, motor domain, electronic control domain or passenger cabin domain, N_switch represents the number of actuator switching, T_response represents the temperature response time, w1 represents the first weight coefficient, w2 represents the second weight coefficient, w3 represents the third weight coefficient, and w4 represents the fourth weight coefficient.

[0223] In yet another specific implementation, the navigation condition identification model is implemented in at least the following ways:

[0224] Obtain route information for the next 5km-10km from the vehicle navigation system and classify it according to operating conditions.

[0225] Working condition classification algorithm:

[0226] Step 1: Obtain the navigation path point set.

[0227] Path={(lat_i,lon_i,dist_i,v_limit_i),i=1...N};

[0228] lat_i, lon_i: latitude and longitude of the i-th path point;

[0229] dist_i: Distance from the current position, in meters (m);

[0230] v_limit_i: Speed ​​limit for the road segment, in km / h.

[0231] Step 2: Terrain feature extraction.

[0232] foriin1toN-1:

[0233] Δh_i=elevation(lat_i+1,lon_i+1)-elevation(lat_i,lon_i)

[0234] grade_i = arctan(Δh_i / dist_i) × 100% # Slope percentage.

[0235] Road classification:

[0236] if|grade_i|<2%:road_type="level road";

[0237] elifgrade_i>=2%:road_type="uphill", slope_level=grade_i;

[0238] elifgrade_i<=-2%:road_type="downhill", slope_level=|grade_i|.

[0239] Step 3: Traffic condition prediction.

[0240] Obtained from real-time traffic data:

[0241] traffic_flow_i: Traffic flow density, ranging from 0 to 1 (0 for smooth flow, 1 for congestion);

[0242] v_actual_i: Actual average vehicle speed, in km / h.

[0243] iftraffic_flow_i<0.3andv_actual_i>0.8×v_limit_i:

[0244] traffic_condition="high-speed cruise";

[0245] eliftraffic_flow_i>0.7orv_actual_i<0.3×v_limit_i:

[0246] traffic_condition="congestion, stop and go";

[0247] else:

[0248] traffic_condition="city roads".

[0249] In yet another specific implementation, the power demand forecasting model is achieved at least in the following ways:

[0250] Future power demand curves can be predicted based on road information:

[0251] Power prediction formula:

[0252] P_predict(t)=P_accel(t)+P_grade(t)+P_resist(t)+P_aux(t);

[0253] In the above formula, P_accel(t) represents the acceleration power, in kW; P_accel = 0.5mv(t)a(t) / 1000, where m represents the total mass of the vehicle, in kg (including load), v(t) represents the predicted vehicle speed, in m / s, and a(t) represents the predicted acceleration, in m / s². 2 .

[0254] Acceleration prediction (can be based on traffic conditions):

[0255] iftraffic_condition=="High-speed cruising":

[0256] a(t) = 0 (uniform speed);

[0257] eliftraffic_condition=="Congestion, Stop & Go":

[0258] a(t) = ±1.5 m / s 2 (Frequent acceleration and deceleration);

[0259] eliftraffic_condition=="City Road":

[0260] a(t) = ±0.8 m / s 2 (Appropriate acceleration and deceleration).

[0261] Furthermore, P_grade(t) represents the ramp power, in kW.

[0262] The calculation method for climbing efficiency is at least as follows:

[0263] P_grade=mgsinαv(t) / 1000;

[0264] In the above formula, g represents the acceleration due to gravity, which is approximately 9.8 m / s². 2 α represents the slope angle, α = arctan(grade_i).

[0265] For example, for an 18-ton vehicle, a 5% gradient, and a driving speed of 40 km / h:

[0266] P_grade=18000×9.8×0.05×(40 / 3.6) / 1000=98kW.

[0267] Furthermore, P_resist(t) represents the driving resistance power, in kW.

[0268] The calculation method for driving resistance power is at least as follows:

[0269] P_resist=(C_rrmg+0.5ρC_dAv 2 v / 1000;

[0270] In the above formula, C_rr represents the rolling resistance coefficient, typically ranging from 0.008 to 0.012; ρ represents the air density, for example, 1.225 kg / m³. 3 C_d represents the drag coefficient, typically 0.6-0.8 for commercial vehicles; A represents the frontal area, in m². 2 .

[0271] Furthermore, P_aux(t) represents the auxiliary system power, in kW.

[0272] The calculation method for auxiliary system power is at least as follows:

[0273] P_aux=P_steer+P_compress+P_24V;

[0274] In the above formula, P_steer represents the power of the power steering, which can be taken as 1kW-4kW depending on the actual vehicle; P_compress represents the power of the air pressure system, which can be taken as 0.5kW-1.5kW depending on the actual vehicle; P_24V represents the power of the 24V electrical equipment, which can be taken as 0.5kW-1kW depending on the actual vehicle.

[0275] In yet another specific implementation, the environmental change prediction model is achieved at least in the following ways:

[0276] It can combine weather forecasts and time information to predict changes in ambient temperature.

[0277] Ambient temperature prediction model:

[0278] T_amb_predict(t)=T_amb_base(t)+ΔT_weather(t)+ΔT_altitude(t);

[0279] In the above formula, T_amb_base(t) represents the base ambient temperature in °C, which can be obtained from the weather API for hourly forecast temperatures.

[0280] Furthermore, ΔT_weather(t) represents the weather effect correction, in °C.

[0281] ifweather=="Sunny":

[0282] ΔT_weather = +2℃ (warming due to solar radiation);

[0283] elifweather=="Cloudy":

[0284] ΔT_weather=0℃;

[0285] elifweather=="rainy day":

[0286] ΔT_weather = -1℃ (evaporative cooling);

[0287] elifweather=="snowy day":

[0288] ΔT_weather=-2℃.

[0289] Furthermore, ΔT_altitude(t) represents the altitude effect correction, in °C.

[0290] The calculation method for altitude effect correction is at least as follows:

[0291] ΔT_altitude=-0.65×(h_predict-h_current) / 100;

[0292] In other words, the temperature drops by 0.65℃ for every 100m increase in altitude.

[0293] Furthermore, the solar radiation intensity is predicted to be at least as follows:

[0294] Q_solar_predict(t)=Q_solar_max×sin(θ_sun)×k_weather;

[0295] In the above formula, Q_solar_max represents the maximum radiation intensity, which can be taken as 1000W / m. 2 (Sunny day at noon).

[0296] Furthermore, θ_sun represents the solar altitude angle, which is related to time and latitude.

[0297] The solar altitude angle can be calculated in at least the following ways:

[0298] θ_sun=arcsin[sin(δ)sin(φ)+cos(δ)cos(φ)cos(h)];

[0299] In the above formula, δ represents the solar declination angle, φ represents the local latitude, h represents the hour angle, and h = 15° × (t - 12).

[0300] Furthermore, k_weather represents the weather coefficient.

[0301] For example,

[0302] Sunny: k_weather=1.0;

[0303] Cloudy: k_weather=0.6;

[0304] Cloudy: k_weather=0.3;

[0305] Rain / snow: k_weather=0.1.

[0306] In another specific implementation, the constraints of the multi-objective optimization function for the whole vehicle are achieved at least in the following ways:

[0307] 1. Temperature constraint:

[0308] T_min≤T_domain≤T_max,∀domain∈{battery, motor, electronic control, cabin}.

[0309] Specific constraints:

[0310] (1) Battery domain: -15℃≤T_bat≤50℃;

[0311] Note: Exceeding this range may result in decreased battery performance or safety risks;

[0312] Optimal range: 15℃-35℃, within which battery performance is best.

[0313] (2) Motor domain: -40℃≤T_mot≤130℃;

[0314] Note: Stator winding temperature rating is H, with a maximum allowable temperature of 130℃;

[0315] Optimal range: 60℃-90℃, highest efficiency and longest lifespan.

[0316] (3) Electrical control domain: -40℃≤T_inv≤150℃;

[0317] Note: IGBT junction temperature limit; exceeding 150℃ triggers over-temperature protection.

[0318] Optimal range: 70℃-110℃, the balance point between switching loss and conduction loss.

[0319] (4) Cabin area: -40℃≤T_cab≤60℃;

[0320] Note: Requirements for passenger comfort and safety;

[0321] Comfort range: 18℃-28℃, adjustable according to the season.

[0322] 2. Power Constraints:

[0323] P_cooling≤P_available;

[0324] In the above formula, P_cooling represents the total power requirement of the cooling system, in kW.

[0325] The total power requirement of the cooling system should be calculated at least in the following ways:

[0326] P_cooling=P_comp+P_pump+P_fan.

[0327] P_available represents the upper limit of available power, in kW, which is limited by the battery output capacity and the power distribution of the whole vehicle.

[0328] The upper limit of available power should be calculated at least in the following ways:

[0329] P_available=min{P_bat_max,P_total_limit}

[0330] In the above formula, P_bat_max represents the maximum discharge power of the battery, which is related to the state of charge (SOC).

[0331] For example,

[0332] SOC>80%: P_available=15kW;

[0333] 50% <SOC≤80%:P_available=12kW;

[0334] 20% <SOC≤50%:P_available=8kW;

[0335] SOC≤20%: P_available=5kW (low power limits cooling power).

[0336] Furthermore, P_total_limit represents the vehicle's power allocation limit, prioritizing power demand.

[0337] In addition, the following sub-constraints:

[0338] 0≤P_comp≤8kW (compressor rated power);

[0339] 0≤P_pump≤2kW (maximum power of water pump);

[0340] 0≤P_fan≤1.5kW (maximum fan power);

[0341] 0≤P_PTC≤6kW (PTC heater power, for winter use only).

[0342] 3. Flow constraints:

[0343] Q_flow_min≤Q_flow≤Q_flow_max.

[0344] Traffic range for each domain:

[0345] Battery coolant flow rate: 20L / min≤Q_bat≤100L / min;

[0346] Q_min guarantees the minimum heat exchange, while Q_max is limited by the pump capacity.

[0347] Motor coolant flow rate: 15L / min≤Q_mot≤80L / min;

[0348] At low speeds, the flow rate can be reduced to lower pump power consumption.

[0349] Electronically controlled coolant flow rate: 10L / min ≤ Q_inv ≤ 60L / min;

[0350] IGBT heatsinks have relatively low flow rate requirements.

[0351] Total flow constraint: Q_bat + Q_mot + Q_inv ≤ 200 L / min;

[0352] Limited by the total flow capacity of the water pump.

[0353] 4. Rate of change constraint:

[0354] |dT / dt|≤Rate_max.

[0355] Temperature change rate limits for each domain:

[0356] Battery: |dT_bat / dt|≤3℃ / min;

[0357] Note: Prevents battery thermal shock and extends battery life.

[0358] Motor: |dT_mot / dt|≤5℃ / min;

[0359] Note: Rapid changes in winding temperature can affect insulation performance.

[0360] Electrical control: |dT_inv / dt|≤8℃ / min;

[0361] Note: Risk of thermal expansion and contraction mismatch between IGBT chip and heat sink.

[0362] Cabin: |dT_cab / dt|≤2℃ / min;

[0363] Note: Passenger comfort requirements; discomfort caused by sudden temperature changes.

[0364] The rate of change constraint can be treated as a soft constraint: a penalty term is added when it is exceeded, but it is not strictly prohibited.

[0365] In another specific implementation, the multi-objective optimization function for the whole vehicle can be specifically defined as follows:

[0366] J=w1E_total+w2ΔT_max+w3N_switch+w4T_response;

[0367] In the above formula, J is dimensionless, and the smaller the value, the better the overall performance.

[0368] Furthermore, E_total represents the total energy consumption of the system, in kWh, which can refer to the total energy consumption of all actuators within the optimization cycle.

[0369] Specifically, the total energy consumption of the system can be calculated at least in the following ways:

[0370] E_total=∫(P_comp+P_pump+P_fan+P_PTC)dt;

[0371] In the above formula, P_comp represents the compressor power, in kW, and can range from 0kW to 8kW; P_pump represents the water pump power, in kW, and can range from 0kW to 2kW; P_fan represents the fan power, in kW, and can range from 0kW to 1.5kW; P_PTC represents the PTC heater power, in kW, and can range from 0kW to 6kW.

[0372] Furthermore, ΔT_max represents the maximum temperature difference in each domain, in °C, which reflects the uniformity of temperature control.

[0373] Specifically, the maximum temperature difference in each region can be calculated at least in the following ways:

[0374] ΔT_max=max{|T_i-T_target_i|,i=1,2,3,4};

[0375] In the above formula, i=1 represents the battery domain, and T_target_1=25℃ (optimal operating temperature).

[0376] i=2 indicates the motor domain, and T_target_2=60℃-80℃ (rated operating temperature).

[0377] i=3 indicates the electronic control domain, and T_target_3=70℃-90℃ (the optimal operating temperature of IGBT).

[0378] i=4 represents the cabin domain, and T_target_4=T_set (i.e., the user-set temperature).

[0379] Furthermore, N_switch represents the number of executor switches, which is dimensionless and can refer to the total number of executor actions within the optimization cycle.

[0380] Specifically, the number of actuator switching times is calculated at least in the following ways:

[0381] N_switch=Σ|u_j(k)-u_j(k-1)|;

[0382] In the above formula, u_j(k) represents the control command of the j-th actuator at time k (normalized to [0,1]). It is understandable that the purpose of setting this parameter is to reduce frequent actuator movements, extend service life, and reduce noise.

[0383] Furthermore, T_response represents the temperature response time, in seconds (s), which can refer to the time from the occurrence of a temperature deviation to the return to the target value.

[0384] Temperature response time can be calculated at least in the following ways:

[0385] T_response=Σ[τ_i×Priority_i];

[0386] In the above formula, τ_i represents the temperature response time constant of the i-th domain; Priority_i represents the response priority coefficient of the i-th domain (high priority domains require fast response, and this parameter can be configured according to the actual vehicle adaptability).

[0387] It can be seen that w1, w2, w3, and w4 are the weight coefficients, which are dimensionless and satisfy Σw_i=1 and w_i≥0.

[0388] For example, the weighting coefficients can be dynamically adjusted according to the operating conditions:

[0389] High-speed operation: w1↑ (energy consumption priority), w1=0.4, w2=0.3, w3=0.15, w4=0.15;

[0390] Urban operating conditions: w2↑ (temperature control priority), w1=0.3, w2=0.4, w3=0.15, w4=0.15;

[0391] Comfort condition: w4↑ (response priority), w1=0.25, w2=0.25, w3=0.15, w4=0.35;

[0392] Life protection: w3↑ (reduce operation), w1=0.3, w2=0.3, w3=0.3, w4=0.1.

[0393] S3. Perform coordinated operation of vehicle thermal management based on each heat load prediction model, navigation condition identification model, power demand prediction model, environmental change prediction model, vehicle multi-objective optimization function and its constraints.

[0394] Among them, the coordinated operation of vehicle thermal management can be implemented based on a variety of strategies.

[0395] Strategy 1: Heat transfer and utilization.

[0396] 1. Criteria for determining waste heat recovery:

[0397] IF(T_motor>50℃)AND(T_cabin_target-T_cabin_actual>5℃)THEN

[0398] Activate the motor-cabin heat exchanger;

[0399] Valve opening degree = min(100%, K_heat*(T_cabin_target-T_cabin_actual));

[0400] K_heat=20% / ℃ / / proportional coefficient.

[0401] 2. Utilization of waste heat from battery preheating:

[0402] IF(T_battery<15℃)AND((T_motor>50℃)OR(T_inverter>50℃))THEN

[0403] Priority judgment:

[0404] IFT_motor-T_battery>30℃THEN

[0405] Open the motor-battery heat exchange bypass;

[0406] Coolant flow rate = max(50%, min(100%, (T_motor-50) / 50*100%));

[0407] ELSEIFT_inverter-T_battery>25℃THEN

[0408] Open the electronic control-battery heat exchange bypass;

[0409] Coolant flow rate = max(40%, min(100%, (T_inverter-45) / 45*100%)).

[0410] 3. Multi-domain heat cascade utilization:

[0411] Heat flow direction decision:

[0412] (1) Sorting of high temperature domains: in descending order of T_motor, T_inverter, T_battery;

[0413] (2) Sort by hot domain: in ascending order of T_target-T_actual;

[0414] (3) Establish the heat flow topology:

[0415] FOReach high temperature domain i:

[0416] FOReach requires hot domain j:

[0417] IF(T_i-T_j>20℃) AND (path available):

[0418] Calculate the heat transfer Q_ij = h_ij * A_ij * (T_i - T_j);

[0419] IFQ_ij>Q_threshold:

[0420] Add to the heat flow list.

[0421] Strategy 2: Dynamic allocation of cooling resources.

[0422] 1. Flow allocation algorithm (based on temperature margin and priority):

[0423] Calculate the temperature margin for each domain:

[0424] Margin_motor=(T_motor_max-T_motor) / T_motor_max;

[0425] Margin_inverter=(T_inverter_max-T_inverter) / T_inverter_max;

[0426] Margin_battery = (T_battery_max - T_battery) / T_battery_max。

[0427] Calculate the normalized demand:

[0428] Demand_motor = (1 - Margin_motor) * Priority_motor;

[0429] Demand_inverter = (1 - Margin_inverter) * Priority_inverter;

[0430] Demand_battery = (1 - Margin_battery) * Priority_battery。

[0431] Total demand:

[0432] Demand_total = Demand_motor + Demand_inverter + Demand_battery。

[0433] Flow distribution:

[0434] Flow_motor = Flow_total * (Demand_motor / Demand_total);

[0435] Flow_inverter = Flow_total * (Demand_inverter / Demand_total);

[0436] Flow_battery = Flow_total * (Demand_battery / Demand_total)。

[0437] Constraint check:

[0438] IF Flow_motor < Flow_motor_min THEN Flow_motor = Flow_motor_min;

[0439] IF Flow_inverter < Flow_inverter_min THEN Flow_inverter = Flow_inverter_min;

[0440] IFFlow_battery <Flow_battery_minTHENFlow_battery=Flow_battery_min。

[0441] 2. Intelligent power allocation for the compressor:

[0442] Calculate the required compressor power for each domain:

[0443] P_comp_motor=Q_motor / COP_motor;

[0444] P_comp_inverter=Q_inverter / COP_inverter;

[0445] P_comp_battery=Q_battery / COP_battery;

[0446] P_comp_cabin=Q_cabin / COP_cabin.

[0447] Total power demand:

[0448] P_comp_demand=P_comp_motor+P_comp_inverter+P_comp_battery+P_comp_cabin;

[0449] IFP_comp_demand>P_comp_maxTHEN

[0450] / / Reduction by priority

[0451] WHILEP_comp_demand>P_comp_max:

[0452] Find the lowest priority field i;

[0453] IFQ_i>Q_i_minTHEN

[0454] Q_i=Q_i*0.95 / / Reduce 5%;

[0455] Recalculate P_comp_demand;

[0456] ELSE

[0457] Skip this field and check the next lowest priority field;

[0458] / / Allocate actual compressor power

[0459] P_comp_actual_motor=min(P_comp_motor,P_comp_max*Demand_motor / Demand_total);

[0460] P_comp_actual_inverter=min(P_comp_inverter,P_comp_max*Demand_inverter / Demand_total);

[0461] P_comp_actual_battery=min(P_comp_battery,P_comp_max*Demand_battery / Demand_total).

[0462] Strategy 3: Predictive temperature control.

[0463] 1. Feedforward control based on road condition prediction:

[0464] Prediction time domain: T_horizon=30s / / 30 seconds from now;

[0465] / / Slope Prediction

[0466] IF detects uphill ahead (slope > 3%, length > 200m) THEN

[0467] Estimated power increment:

[0468] ΔP_motor=m*g*sin(θ)*v / / Additional power to overcome gravity;

[0469] ΔP_inverter = ΔP_motor * 0.1 / / The power of the electronic control increases accordingly;

[0470] Forecast temperature rise:

[0471] ΔT_motor_predict=ΔP_motor*t_slope / (C_motor*m_motor);

[0472] ΔT_inverter_predict=ΔP_inverter*t_slope / (C_inverter*m_inverter);

[0473] Pre-cooling control:

[0474] T_motor_target_precooling=T_motor_target-ΔT_motor_predict*0.8;

[0475] T_inverter_target_precooling=T_inverter_target-ΔT_inverter_predict*0.8;

[0476] Pre-cooling should begin 10 seconds in advance:

[0477] Increase compressor speed: n_comp = n_comp_normal * 1.2;

[0478] Increase coolant flow rate: Flow = Flow_normal * 1.15;

[0479] / / Downhill prediction (energy recovery)

[0480] IF detects downhill ahead (slope < -3%, length > 200m) THEN

[0481] Estimated recovery power:

[0482] P_regen=min(P_regen_max,m*g*sin(|θ|)*v*η_regen);

[0483] Estimated battery temperature rise:

[0484] ΔT_battery_predict=P_regen*t_slope / (C_battery*m_battery);

[0485] Battery preheating control (winter):

[0486] IFT_ambient<10℃ANDT_battery<20℃THEN

[0487] T_battery_target_preheat=25℃ / / Raise to the optimal operating temperature;

[0488] Start warm-up 15 seconds in advance.

[0489] 2. Calculation of dynamically adjusted lead time:

[0490] Calculate the lead time based on vehicle speed and predicted distance:

[0491] t_lead=d_predict / v;

[0492] In the above formula, d_predict represents the predicted event distance (meters), v represents the current vehicle speed (m / s), and t_lead represents the lead time (seconds).

[0493] Lead time adjustment:

[0494] IFt_lead<5sTHEN

[0495] / / Time is too short, increase response strength

[0496] Response intensity coefficient = 1.5;

[0497] ELSEIFt_lead>20sTHEN

[0498] / / Ample time, gradual adjustment

[0499] Response intensity coefficient = 0.8;

[0500] ELSE

[0501] Response intensity coefficient = 1.0.

[0502] Furthermore, the coordinated operation of vehicle thermal management can be controlled and executed in real time.

[0503] 1. Control command generation:

[0504] (1) Compressor speed control:

[0505] n_comp=f(Q_cooling_total,T_evap,T_cond);

[0506] In the above formula, n_comp represents the compressor speed command, in rpm, and the range can be 1000rpm-6000rpm.

[0507] Q_cooling_total represents the total cooling requirement, in kW.

[0508] Total cooling requirements can be calculated at least in the following ways:

[0509] Q_cooling_total=ΣQ_i;

[0510] In the above formula, Q_i represents the cooling demand of each domain, which can be obtained from the heat load prediction model in step S2 above; T_evap represents the evaporator temperature in °C, which affects the cooling efficiency; T_cond represents the condenser temperature in °C, which is affected by the ambient temperature.

[0511] Furthermore, the calculation method is provided:

[0512] Base speed: n_base=k1×Q_cooling_total, where k1 is a proportional coefficient (e.g., 200rpm / kW);

[0513] Temperature correction: n_comp = n_base[1 + k2(T_cond - T_evap - ΔT_nominal)];

[0514] In the above formula, ΔT_nominal represents the nominal temperature difference, typically 30℃; k2 represents the correction factor, typically 0.02 / ℃.

[0515] Range constraint: n_comp = max(1000, min(n_comp, 6000));

[0516] Rate of change limit: |dn_comp / dt|≤500rpm / s, to avoid compressor shock.

[0517] (2) Pump flow control:

[0518] Q_pump_i=f(Q_cooling_i,Priority_i,T_margin_i);

[0519] In the above formula, Q_pump_i represents the coolant flow rate allocated to the i-th domain, in L / min; Q_cooling_i represents the cooling demand of the i-th domain, in kW; and Priority_i represents the priority coefficient of the i-th domain, dimensionless, ranging from 0 to 1.

[0520] Specifically, Priority_bat=1.0 (highest battery priority, safety);

[0521] Priority_mot=0.8 (motor is the next priority, performance is the next priority);

[0522] Priority_inv=0.7 (Electrical control is third, reliability);

[0523] Priority_cab=0.5 (cabin minimum, comfort).

[0524] Furthermore, T_margin_i represents the temperature margin of the i-th domain, in °C.

[0525] The temperature margin of the i-th domain can be determined at least by the following means:

[0526] T_margin_i=T_max_i-T_current_i;

[0527] The smaller the temperature margin (the closer to the upper limit), the more flow rate is allocated.

[0528] Furthermore, the following calculation method (dynamic allocation algorithm) is provided:

[0529] Step 1: Calculate the basic flow requirements.

[0530] Q_base_i = k_flow·Q_cooling_i, where k_flow represents the heat transfer coefficient, which can be taken as 5 L / (min·kW).

[0531] Step 2: Adjust according to priority and temperature margin.

[0532] Weight_i=Priority_i / (T_margin_i+ε);

[0533] ε is a small quantity to prevent division by zero; a typical value can be taken as 1℃.

[0534] Step 3: Normalize the allocation of total flow.

[0535] Q_pump_i=Q_base_i×Weight_i / ΣWeight_j×Q_total_available;

[0536] In the above formula, Q_total_available represents the total flow capacity of the water pump, for example, 200L / min.

[0537] Step 4: Apply traffic constraints.

[0538] Q_pump_i=max(Q_min_i,min(Q_pump_i,Q_max_i)).

[0539] (3) Fan speed control:

[0540] n_fan_i=f(T_radiator_i,v_vehicle,T_amb);

[0541] In the above formula, n_fan_i represents the speed command of the i-th fan, in rpm, and can be in the range of 0rpm-3000rpm; T_radiator_i represents the temperature of the i-th radiator, in °C; v_vehicle represents the vehicle speed, in km / h, which affects the natural wind speed; T_amb represents the ambient temperature, in °C.

[0542] Furthermore, the following calculation method is provided:

[0543] Step 1: Calculate the required wind speed.

[0544] v_air_need=k_rad×(T_radiator-T_amb);

[0545] In the above formula, k_rad represents the heat transfer coefficient of the radiator, and a typical value can be taken as 5 (km / h) / ℃.

[0546] Step 2: Compensate for the impact of vehicle speed.

[0547] v_air_net=max(0,v_air_need-0.5v_vehicle);

[0548] When driving at high speeds, the natural wind speed is high, which can reduce the fan speed.

[0549] Step 3: Rotation speed calculation.

[0550] n_fan = k_fan × v_air_net;

[0551] In the above formula, k_fan represents the fan characteristic coefficient, which can be taken as 150rpm / (km / h).

[0552] Step 4: Range limitation and dead zone handling.

[0553] if n_fan < 500: n_fan = 0 (to avoid inefficient work);

[0554] n_fan=min(n_fan,3000) (Maximum speed limit).

[0555] Step 5: Hysteresis control avoids frequent start-stop cycles.

[0556] Startup threshold: T_radiator > T_target + 5℃;

[0557] Stop threshold: T_radiator <T_target+2℃;

[0558] (4) Valve opening control:

[0559] θ_valve_ij=f(Heat_transfer_ij,Q_flow_i,Q_flow_j);

[0560] In the above formula, θ_valve_ij represents the valve opening degree connecting domain i and domain j, which is a percentage and can be 0-100%.

[0561] Heat_transfer_ij represents the heat transfer requirement from domain i to domain j, in kW.

[0562] The heat transfer requirement is positive: heat is transferred from i to j.

[0563] The heat transfer requirement is negative: heat is transferred from j to i.

[0564] Q_flow_i and Q_flow_j represent the flow rates of domain i and domain j, respectively, in L / min.

[0565] Furthermore, the following calculation method (heat transfer control) is provided:

[0566] Step 1: Determine the direction and feasibility of heat transfer.

[0567] if(T_i>T_j)and(Heat_transfer_ij>0):

[0568] The valve can be opened to allow heat to flow from i to j;

[0569] else:

[0570] The valve is closed to prevent reverse heat transfer.

[0571] Step 2: Calculate the required valve opening.

[0572] ΔT_design = 10℃ (design temperature difference);

[0573] Q_max_transfer=c_p×ρ×Q_flow×ΔT_design;

[0574] In the above formula, c_p represents the specific heat capacity, for example, it can be taken as 4.2kJ / (kg×℃); ρ represents the density, for example, it can be taken as 1.0kg / L.

[0575] θ_valve=(Heat_transfer_ij / Q_max_transfer)×100%.

[0576] Step 3: Scope limitation.

[0577] θ_valve=max(0,min(θ_valve,100)).

[0578] Step 4: Minimum opening constraint (to avoid throttling losses).

[0579] if0<θ_valve<15%:θ_valve=15%.

[0580] For example, a typical valve configuration:

[0581] ①θ_valve_bat-mot: Battery-motor heat exchange valve (preheats the battery with waste heat from the motor in winter);

[0582] ②θ_valve_mot-cab: Motor-cabin heat exchange valve (waste heat heating in winter);

[0583] ③θ_valve_inv-cab: Electrically controlled cabin heat exchange valve (waste heat utilization in winter).

[0584] 2. Actuator Coordination Control:

[0585] (1) Establish an executor priority queue;

[0586] (2) Avoid frequent actuator switching;

[0587] (3) Achieve smooth transition control.

[0588] To avoid resource contention, frequent switching, or mutual interference among actuators during multi-domain collaborative control, this invention introduces an actuator priority queue in actuator coordination control. The principle and basis for its establishment are as follows:

[0589] The principle of prioritizing safety should first and foremost follow the safety requirements of the entire vehicle and its critical components.

[0590] Actuators that directly affect the overheating / overcooling risk of key components such as power batteries and power devices (such as cooling pumps, key valves, and compressors) are given the highest priority.

[0591] Actuators that only affect occupant comfort or minor energy consumption optimization (such as some branch valves or fan speed fine-tuning) have relatively low priority. This principle is determined based on vehicle thermal runaway protection, power device temperature ratings, and other safety specifications and vehicle calibration experience.

[0592] The energy consumption and benefit trade-off is based on ranking the actuators according to "energy consumption benefit per unit execution cost" under the premise that safety constraints are met:

[0593] Actuators that can significantly reduce system energy consumption or improve thermal management efficiency (such as cooling pump speed and compressor speed) have higher priority in the queue;

[0594] Actuators with limited energy benefits or indirect effects (such as local airflow fine-tuning) should have their priority appropriately reduced. The above ranking is based on bench tests and vehicle simulation results, and is obtained by evaluating the "energy consumption change / control action amplitude" of different actuators under typical operating conditions.

[0595] Dynamic response characteristics are based on consideration of the response speed and dynamic characteristics of each actuator:

[0596] Actuators with fast response and significant impact on system dynamics (such as electronic valves and electronic water pumps) should be prioritized for use when operating conditions change rapidly.

[0597] For actuators with slow response or significant inertial hysteresis (such as certain mechanical valves and slowly changing fan systems), frequent start-stop cycles are avoided by reducing their priority and introducing a hysteresis region. This design is determined based on the actuator's physical characteristics and the manufacturer's technical parameters, combined with the control system's dynamic performance requirements.

[0598] The constraints on action frequency and lifespan are based on extending actuator lifespan and reducing maintenance costs. This invention introduces an "action frequency penalty factor" into the priority queue:

[0599] For actuators that are prone to fatigue or wear due to frequent operations, their priority will be dynamically reduced when they are scheduled multiple times in a short period of time, prompting the system to prioritize achieving the goal through other actuators.

[0600] For actuators with simple structures and low operating costs, a high and relatively stable priority can be maintained. This mechanism is designed based on the actuator's lifespan curve, maintenance strategy, and overall vehicle reliability requirements.

[0601] The multi-objective comprehensive scoring mechanism integrates multiple dimensions such as safety, energy consumption benefit, dynamic response, and lifespan constraints. In this embodiment, a weighted scoring model is designed for each actuator:

[0602] Score_k=w_s\cdotS_k+w_e\cdotE_k+w_d\cdotD_k+w_l\cdotL_k;

[0603] In the above formula, S_k is the safety importance score, E_k is the energy consumption benefit score, D_k is the dynamic response score, L_k is the lifespan / maintenance sensitivity score, and w_s, w_e, w_d, and w_l are the corresponding weight coefficients. Actuators are sorted from highest to lowest according to Score_k, forming an actuator priority queue. The weight coefficients can be determined through calibration and simulation optimization based on different vehicle models and application scenarios to ensure optimal overall safety, energy consumption, and comfort across different products.

[0604] The technical solution provided in this embodiment firstly acquires at least the battery domain state information, motor domain state information, electronic control domain state information, passenger cabin domain state information, and overall vehicle operating status information. Further, based on a preset algorithm architecture and each state information, it establishes at least the following models: battery domain thermal load prediction model, motor domain thermal load prediction model, electronic control domain thermal load prediction model, passenger cabin thermal load prediction model, navigation condition identification model, power demand prediction model, and environmental change prediction model. A multi-objective optimization function for the entire vehicle and its constraints are then constructed based on the state information. Finally, a coordinated operation for overall vehicle thermal management is performed based on each thermal load prediction model, navigation condition identification model, power demand prediction model, environmental change prediction model, and the multi-objective optimization function for the entire vehicle and its constraints. Therefore, this embodiment can at least achieve coordinated control of the four major thermal management domains—battery, motor, electronic control, and passenger cabin—in new energy commercial vehicles. This ensures the accuracy of temperature control in each domain while achieving multi-objective optimization of overall vehicle energy efficiency, system lifespan, and response speed, thus improving the user's driving experience.

[0605] It should be noted that, unlike the aforementioned heat load prediction models, in another specific implementation, the battery domain heat load prediction model is achieved in at least the following ways:

[0606] Q_bat=f(I,SOC,T_bat,T_amb);

[0607] In the above formula, SOC represents the battery's state of charge, as a percentage, ranging from 0 to 100%.

[0608] In yet another specific implementation, the motor domain heat load prediction model is achieved at least in the following ways:

[0609] Q_mot=f(P_mot,n,T_mot);

[0610] In the above formula, n represents the motor speed, in rpm, ranging from 0 to 10000 rpm; T_mot represents the motor stator temperature, in °C, with an operating range of -40 °C to 150 °C.

[0611] In yet another specific implementation, the electrical control domain heat load prediction model is achieved at least in the following ways:

[0612] Q_inv=f(P_inv,f_sw,T_inv);

[0613] In the above formula, T_inv represents the IGBT junction temperature in °C, with an operating range of -40°C to 150°C and a limit of 175°C.

[0614] In yet another specific implementation, the cabin area heat load prediction model is achieved at least through the following means:

[0615] Q_cab=f(T_set,T_cab,T_amb,Q_solar);

[0616] In the above formula, Q_cab represents the cabin heat load in W, with a positive value indicating that cooling is required and a negative value indicating that heating is required; T_set represents the cabin temperature setpoint in °C, which can be set by the user in the range of 18°C-28°C.

[0617] Figure 2 This is a schematic diagram of a vehicle thermal management device provided in an embodiment of the present invention. This embodiment is applicable to at least any integrated thermal management scenario for new energy medium and heavy-duty commercial vehicles. The vehicle thermal management device can be implemented using software and / or hardware. Figure 2 As shown, the vehicle thermal management device includes at least:

[0618] The information acquisition module 110 is used to acquire at least the vehicle's battery domain status information, motor domain status information, electronic control domain status information, passenger cabin domain status information, and overall vehicle operating status information.

[0619] The model building module 120 is used to build at least the following models based on the preset algorithm architecture and each state information: battery domain thermal load prediction model, motor domain thermal load prediction model, electronic control domain thermal load prediction model, cabin domain thermal load prediction model, navigation condition identification model, power demand prediction model and environmental change prediction model. It also constructs the vehicle multi-objective optimization function and its constraints based on the state information.

[0620] The collaborative management module 130 is used to perform collaborative operations for vehicle thermal management based on each thermal load prediction model, navigation condition identification model, power demand prediction model, environmental change prediction model, vehicle multi-objective optimization function and its constraints.

[0621] Optionally, the battery domain heat load prediction model can be implemented at least in the following way: Q_bat=I 2 R_int+IVη_loss+k_conv(T_bat-T_amb); In the above formula, Q_bat represents the battery thermal load, I 2 R_int represents the heat generated by the battery's internal resistance, I represents the battery's charging and discharging current, R_int represents the battery's equivalent internal resistance, IVη_loss represents the battery's charging and discharging loss, V represents the battery's charging and discharging voltage, η_loss represents the battery's loss coefficient, k_conv(T_bat-T_amb) represents the battery's convective heat dissipation, k_conv represents the battery's convection coefficient, T_bat represents the battery's average temperature, and T_amb represents the ambient temperature.

[0622] In yet another specific implementation, the motor domain heat load prediction model can optionally be implemented at least in the following way: Q_mot=P_mot(1-η_mot)+P_iron+P_mech;

[0623] In the above formula, Q_mot represents the motor thermal load, P_mot(1-η_mot) represents the motor copper loss and iron loss, P_mot represents the motor output power, η_mot represents the motor efficiency, P_iron represents the motor iron loss, and P_mech represents the motor mechanical loss.

[0624] In yet another specific implementation, the electronically controlled domain heat load prediction model can optionally be implemented at least in the following way: Q_inv=P_sw+P_cond+P_cap;

[0625] In the above formula, Q_inv represents the electrical control thermal load, P_sw represents the switching loss, P_cond represents the conduction loss, and P_cap represents the capacitor loss.

[0626] In another specific implementation, the cabin area heat load prediction model can optionally be implemented at least in the following way: Q_cab=Q_conv+Q_solar+Q_human+Q_equip; where Q_conv represents convective heat transfer, Q_solar represents solar radiation, Q_human represents passenger heat dissipation, and Q_equip represents equipment heat dissipation.

[0627] Optionally, the multi-objective optimization function for the whole vehicle can be implemented at least in the following way: J = w1E_total + w2ΔT_max + w3N_switch + w4T_response; In the above formula, J represents the multi-objective optimization function for the whole vehicle, E_total represents the total energy consumption of the system, ΔT_max represents the maximum temperature difference in the battery domain, motor domain, electronic control domain or passenger compartment domain, N_switch represents the number of actuator switching, T_response represents the temperature response time, w1 represents the first weight coefficient, w2 represents the second weight coefficient, w3 represents the third weight coefficient, and w4 represents the fourth weight coefficient.

[0628] The technical solution provided in this embodiment firstly acquires at least the vehicle's battery domain state information, motor domain state information, electronic control domain state information, cabin domain state information, and overall vehicle operating status information through an information acquisition module. Further, based on a preset algorithm architecture and each state information, a model building module establishes at least the following models: battery domain thermal load prediction model, motor domain thermal load prediction model, electronic control domain thermal load prediction model, cabin domain thermal load prediction model, navigation condition identification model, power demand prediction model, and environmental change prediction model. A multi-objective optimization function for the entire vehicle and its constraints are then constructed based on the state information. Finally, a collaborative management module performs collaborative thermal management operations for the entire vehicle based on each thermal load prediction model, navigation condition identification model, power demand prediction model, environmental change prediction model, and the multi-objective optimization function for the entire vehicle and its constraints. Therefore, this embodiment can at least achieve collaborative control of the four major thermal management domains—battery, motor, electronic control, and cabin—in new energy commercial vehicles. This ensures the accuracy of temperature control in each domain while achieving multi-objective optimization of vehicle energy efficiency, system lifespan, and response speed, thus enhancing the user's driving experience.

[0629] This embodiment provides an electronic device. Figure 3 This is a schematic diagram of the structure of an electronic device provided in an embodiment of the present invention. See also: Figure 3The electronic device 1000 includes a processor 1001 and a memory 1002. The memory 1002 stores computer-readable instructions. When the computer-readable instructions are executed by the processor 1001, the steps in any of the above-mentioned vehicle thermal management methods are performed. Through the above technical solution, the processor 1001 and the memory 1002 are interconnected and communicate with each other through a communication bus and / or other forms of connection mechanism (not shown). The memory 1002 stores a computer program executable by the processor. When the electronic device 1000 is running, the processor 1001 executes the computer program to execute the vehicle thermal management method in any optional implementation of the above embodiments, so as to achieve at least the following functions: at least obtain the vehicle's battery domain state information, motor domain state information, electronic control domain state information, passenger cabin domain state information, and vehicle operating state information; based on the preset algorithm architecture and each state information, at least establish the vehicle's battery domain thermal load prediction model, motor domain thermal load prediction model, electronic control domain thermal load prediction model, passenger cabin domain thermal load prediction model, navigation condition identification model, power demand prediction model, and environmental change prediction model, and construct the vehicle multi-objective optimization function and its constraints according to the state information; and perform vehicle thermal management collaborative operation according to each thermal load prediction model, navigation condition identification model, power demand prediction model, environmental change prediction model, vehicle multi-objective optimization function, and its constraints.

[0630] This embodiment provides a computer-readable storage medium storing a computer program. When executed by a processor, the program implements the vehicle thermal management method provided in all embodiments of this application: acquiring at least the vehicle's battery domain state information, motor domain state information, electronic control domain state information, passenger cabin domain state information, and vehicle operating state information; establishing at least the vehicle's battery domain thermal load prediction model, motor domain thermal load prediction model, electronic control domain thermal load prediction model, passenger cabin domain thermal load prediction model, navigation condition identification model, power demand prediction model, and environmental change prediction model based on a preset algorithm architecture and each state information; constructing a vehicle multi-objective optimization function and its constraints based on the state information; and performing coordinated vehicle thermal management operations based on each thermal load prediction model, navigation condition identification model, power demand prediction model, environmental change prediction model, vehicle multi-objective optimization function, and its constraints.

[0631] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some or all of the technical features; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present invention.

Claims

1. A vehicle thermal management method, characterized in that, The vehicle thermal management method is applicable at least to new energy commercial vehicles; The vehicle thermal management method includes at least the following: At least the vehicle's battery domain status information, motor domain status information, electronic control domain status information, passenger cabin domain status information, and overall vehicle operating status information should be obtained; Based on the preset algorithm architecture and each state information, at least the following models are established for predicting vehicle thermal load: battery domain thermal load, motor domain thermal load, electronic control domain thermal load, cabin domain thermal load, navigation condition identification, power demand prediction, and environmental change prediction. A multi-objective optimization function for the whole vehicle and its constraints are constructed based on the state information. The vehicle thermal management collaborative operation is performed based on each of the aforementioned heat load prediction model, navigation condition identification model, power demand prediction model, environmental change prediction model, vehicle multi-objective optimization function, and their constraints.

2. The vehicle thermal management method according to claim 1, characterized in that, The battery domain heat load prediction model is implemented in at least the following ways: Q_bat=I 2 R_int+IVη_loss+k_conv(T_bat-T_amb); In the above formula, Q_bat represents the battery thermal load, and I 2 R_int represents the heat generated by the battery's internal resistance, I represents the battery's charging and discharging current, R_int represents the battery's equivalent internal resistance, IVη_loss represents the battery's charging and discharging loss, V represents the battery's charging and discharging voltage, η_loss represents the battery's loss coefficient, k_conv(T_bat-T_amb) represents the battery's convective heat dissipation, k_conv represents the battery's convection coefficient, T_bat represents the battery's average temperature, and T_amb represents the ambient temperature.

3. The vehicle thermal management method according to claim 1, characterized in that, The motor domain heat load prediction model is implemented in at least the following ways: Q_mot=P_mot(1-η_mot)+P_iron+P_mech; In the above formula, Q_mot represents the motor thermal load, P_mot(1-η_mot) represents the motor copper loss and iron loss, P_mot represents the motor output power, η_mot represents the motor efficiency, P_iron represents the motor iron loss, and P_mech represents the motor mechanical loss.

4. The vehicle thermal management method according to claim 1, characterized in that, The electronically controlled domain heat load prediction model is implemented in at least the following ways: Q_inv = P_sw + P_cond + P_cap; In the above formula, Q_inv represents the electrical control thermal load, P_sw represents the switching loss, P_cond represents the conduction loss, and P_cap represents the capacitor loss.

5. The vehicle thermal management method according to claim 1, characterized in that, The cabin area heat load prediction model is achieved through at least the following methods: Q_cab=Q_conv+Q_solar+Q_human+Q_equip; In the above formula, Q_conv represents convective heat transfer, Q_solar represents solar radiation, Q_human represents occupant heat dissipation, and Q_equip represents equipment heat dissipation.

6. The vehicle thermal management method according to claim 1, characterized in that, The preset algorithm architecture employs at least a recurrent neural network incorporating a long short-term memory layer.

7. The vehicle thermal management method according to claim 1, characterized in that, The multi-objective optimization function for the entire vehicle is implemented in at least the following ways: J=w1E_total+w2ΔT_max+w3N_switch+w4T_response; In the above formula, J represents the multi-objective optimization function of the whole vehicle, E_total represents the total energy consumption of the system, ΔT_max represents the maximum temperature difference in the battery domain, motor domain, electronic control domain or cabin domain, N_switch represents the number of actuator switching, T_response represents the temperature response time, w1 represents the first weight coefficient, w2 represents the second weight coefficient, w3 represents the third weight coefficient, and w4 represents the fourth weight coefficient.

8. A vehicle thermal management device, characterized in that, Used to perform the vehicle thermal management method according to any one of claims 1-7; The vehicle thermal management device includes at least: The information acquisition module is used to acquire at least the vehicle's battery domain status information, motor domain status information, electronic control domain status information, passenger cabin domain status information, and overall vehicle operating status information. The model building module is used to build at least the following models based on the preset algorithm architecture and each state information: battery domain thermal load prediction model, motor domain thermal load prediction model, electronic control domain thermal load prediction model, cabin domain thermal load prediction model, navigation condition identification model, power demand prediction model and environmental change prediction model, and to construct the vehicle multi-objective optimization function and its constraints based on the state information. The collaborative management module is used to perform collaborative operations for vehicle thermal management based on each of the aforementioned thermal load prediction models, navigation condition identification models, power demand prediction models, environmental change prediction models, vehicle multi-objective optimization functions, and their constraints.

9. An electronic device comprising a memory and a processor, the memory storing a computer program executable on the processor, characterized in that, When the processor executes the program, it implements the steps in the vehicle thermal management method according to any one of claims 1-7.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by the processor, it implements the steps of the vehicle thermal management method according to any one of claims 1-7.

Images