A method for temperature and torque suppression control of electric heavy-duty truck motors
By combining intelligent control strategies with model predictive control and recurrent neural networks, the motor torque output is dynamically adjusted, solving the problem of temperature and torque balance of electric heavy-duty truck motors under different ambient temperatures and operating conditions, achieving efficient and stable motor operation, and ensuring vehicle safety.
Patent Information
- Application Number
- CN202510586581.X
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-05-08
- Publication Date
- 2026-11-17
AI Technical Summary
Existing electric heavy-duty truck motor temperature control methods cannot achieve a balance between motor torque output and operating time under different ambient temperatures and operating conditions, resulting in large fluctuations in motor performance and failing to meet the requirements for motor performance and safe and reliable operation of the entire vehicle.
By adopting an intelligent control strategy, combining model predictive control using support vector machines and recurrent neural networks, the motor temperature changes are predicted through real-time vehicle data acquisition, and the motor torque output is dynamically adjusted to establish a balance between motor temperature and torque, thereby achieving over-temperature torque limiting control of the motor.
It effectively suppresses motor torque fluctuations, improves motor operating efficiency, extends motor operating time, ensures safe and reliable vehicle operation, and maximizes the utilization of motor efficiency.
Smart Images

Figure CN123104414A_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of electric heavy-duty vehicle technology, and specifically to a method for controlling the temperature and torque of an electric heavy-duty truck motor. Background Technology
[0002] The rise of new energy electric trucks is an important direction for the future development of the heavy truck industry. As a key component of the power system, the motor's performance and reliability must be guaranteed. In harsh working environments such as high ambient temperature, high humidity, and poor ventilation, or when the motor is under prolonged heavy load, insufficient cooling leads to heat accumulation inside the motor, increased internal resistance, and consequently, a decrease in armature back electromotive force, reduced motor torque, or even damage. Therefore, it is necessary to develop reasonable control strategies to protect the motor.
[0003] Therefore, traditional motor temperature control often employs conventional linear or piecewise curves to rigidly and laggingly control motor torque based on motor temperature. When the motor temperature reaches a preset upper limit, the motor output torque is limited to 50% of the initial torque. If reducing the torque still fails to effectively control the motor temperature, the motor output torque is directly reduced to 0. In current electric heavy-duty truck motor temperature and torque control systems, traditional simple linear or piecewise temperature and torque control cannot maximize the utilization of motor output performance or balance motor torque output with operating time. Motor protection control strategies typically involve directly limiting the motor torque to 1 / 2 when the motor temperature reaches a set warning temperature limit. If limiting the torque fails to control the motor temperature, the motor torque is reduced to 0, effectively shutting down the motor. At this point, the vehicle loses power, which can even lead to driving safety issues under certain conditions. This control method, which only considers a single over-temperature torque limitation, cannot achieve dynamic optimal allocation of motor temperature and motor output torque based on real-time environmental and vehicle operating conditions. This not only results in excessive fluctuations in motor torque but also prevents the electric drive system from obtaining continuous and smooth torque, thus failing to meet the comprehensive control requirements for motor performance and safe and reliable vehicle operation.
[0004] As the torque of electric heavy-duty truck motors continues to increase, operating conditions become wider, and operating durations extend, pre-set linear or piecewise curve control cannot meet the requirements for wide-threshold temperature torque control of the motor. Therefore, it is necessary to design a temperature and torque stabilization control method for electric heavy-duty truck motors to solve the problem of over-temperature and torque limiting of existing electric heavy-duty truck motors under different ambient temperatures and operating conditions. Summary of the Invention
[0005] To address the problems existing in the prior art, this invention provides a method for controlling the temperature and torque of an electric heavy-duty truck motor. Based on real-time dynamic analysis and prediction of environmental factors and operating conditions, the method controls the motor torque in a timely manner, maximizing motor efficiency and achieving efficient, stable, and economical operation of the electric heavy-duty truck. By proposing a motor temperature prediction model for the electric heavy-duty truck and designing an intelligent control strategy, combined with a support vector machine model to predict and control the output motor torque threshold, the method achieves over-temperature motor torque control for the electric heavy-duty truck. Within the over-temperature operating range, it ensures maximum motor output torque, preventing the motor from operating at high power under high temperatures, which would affect motor performance and the safe and reliable operation of the electric heavy-duty truck.
[0006] The technical solution adopted by this invention to solve its technical problem is: a method for temperature and torque suppression control of an electric heavy-duty truck motor, comprising the following steps:
[0007] S1: Real-time vehicle operation data is collected using onboard sensors;
[0008] S2: Preprocess the collected data to remove outliers and failure values, and combine data analysis algorithms to predict the motor temperature change trend in real time;
[0009] S3: Based on historical vehicle operation data and neural network algorithms, a motor temperature prediction model is constructed to predict future temperature changes of the motor according to the motor's operating status and external environmental conditions.
[0010] S4: Combining recurrent neural networks to design intelligent control strategies, when the motor temperature is predicted to exceed the safety threshold, the motor output torque is automatically reduced;
[0011] S5: By integrating the motor temperature prediction model with the motor torque control system, the motor torque output is dynamically adjusted to smooth out motor torque fluctuations.
[0012] Preferably, the vehicle operation data collected in step S1 includes, but is not limited to, the ambient temperature T. env Vehicle load M veh Vehicle operating conditions S act Vehicle acceleration V a Driving speed V b Coolant temperature T COOLANT Current I, voltage U.
[0013] Preferably, the motor temperature prediction model in step S2 establishes a balance relationship between motor temperature and motor output torque, and the balance relationship formula is described as follows:
[0014]
[0015] in:
[0016] L - A balance between meeting the motor temperature limit and maximizing the motor output torque;
[0017] T s,t -When the vehicle is operating under condition s, the predicted motor temperature at time t;
[0018] Z s,t -When the vehicle is operating under condition s, the maximum output torque of the motor within the motor temperature limit range at time t;
[0019] λ1, λ2 - Weighting coefficients for motor temperature and motor output torque.
[0020] Preferably, the motor temperature prediction model is based on a combination of ambient temperature T. env Vehicle load M veh Vehicle operating conditions S act Vehicle acceleration V a Driving speed V b Coolant temperature T COOLANT The analysis models for current I and voltage U include, but are not limited to, model predictive control based on vector machines, recurrent neural networks, data-driven models, and physics-driven models. The analysis models are used at different ambient temperatures T. env and vehicle operating conditions S act This provides technical support for the motor torque control strategy.
[0021] Preferably, the state variable x in the model predictive control of the motor temperature prediction model is... k Includes, but is not limited to, current I, voltage U, and control quantity u. k Including but not limited to motor temperature T MOT Disturbance d k Including but not limited to ambient temperature T env Vehicle load M veh Vehicle operating conditions S act Vehicle acceleration V a Driving speed V b Coolant temperature T COOLANT The output includes, but is not limited to, the motor torque threshold Z. T The intermediate quantity m of model predictive control k Including but not limited to motor operating temperature T mot Motor operating power P mot Motor speed N mot ;
[0022] Model predictive control acquires data in real time, including the system's disturbance quantity d. k With state variable x kThe collected data is input into the motor temperature prediction model for calculation. Based on the objective function and constraints, the state variables x are estimated through an optimization solver. k+1 =f(x) k ,u k ,d k ) and optimal control quantity u k The calculation is performed; finally, the calculated control quantity u is... k The first step is to input the motor torque control system, which is used for optimized control of the motor output torque.
[0023] Preferably, the motor temperature prediction model sets constraint one during the model predictive control process: ensuring that the motor operating temperature is within the over-temperature limit range, and the motor temperature T... MOT ≤ Maximum motor temperature T max This allows the motor torque control system to adjust the motor operating temperature Tmot according to actual needs, ensuring that the motor operates within a temperature range below the motor's limit temperature value.
[0024] Preferably, the motor temperature prediction model sets constraint condition two during the model predictive control process: maintaining the actual torque Z of the motor operation. Tmove ≥ Motor torque threshold Z T Battery load P bat ≤ Maximum output power of the battery P bat,max This ensures that the motor can operate reliably and safely when both the predicted operating temperature and the predicted torque threshold of the motor meet the operating requirements.
[0025] Preferably, the objective function is a multi-objective optimization model prediction objective function. The model predictive control strategy defines a suitable objective function minL, which typically includes the motor operating temperature T. mot and meet the motor operating temperature T mot The target is the optimal output torque of the motor, or a combination of both. The optimal output torque of the motor is determined by the ambient temperature T. env and vehicle operating conditions S act Decide.
[0026] Preferably, the recurrent neural network in step S4 serves as the mathematical prediction model for model predictive control, and the neural network receives the current system disturbance d. k State variable x k Control quantity u k Using other intermediate quantities as input, the output is a predicted future system state.
[0027] Preferably, the motor operating temperature T mot With ambient temperature T env Vehicle load M veh Vehicle operating conditions S actThe corresponding relationships are calibrated in advance, and the value of the motor output torque is calculated based on these relationships.
[0028] The present invention has the following beneficial effects:
[0029] The electric heavy-duty truck motor temperature and torque suppression control method designed in this invention dynamically adjusts the motor torque control based on real-time data of factors such as ambient temperature, vehicle load, and road conditions (gradient). This effectively avoids excessive motor temperature and battery overload, maximizing motor operating time while ensuring adequate motor output. Furthermore, this invention employs recurrent neural network technology and model predictive control strategies to improve the system's prediction accuracy and control performance, achieving intelligent and high-precision control of motor torque.
[0030] The electric heavy-duty truck motor temperature and torque stabilization control method designed in this invention dynamically adjusts the motor output torque based on predicted motor temperature using real-time data of ambient temperature and vehicle operating conditions. This effectively suppresses torque fluctuations, improves motor operating efficiency and performance, and meets motor temperature requirements under a wide range of operating conditions, ensuring safe and reliable vehicle operation. Furthermore, this invention combines model predictive control strategies and recurrent neural network technology to improve the prediction accuracy and control effect of the electric drive system, achieving intelligent optimization adjustment of motor output torque at the optimal motor temperature. Attached Figure Description
[0031] Figure 1 This is a model prediction framework diagram for the temperature and torque suppression control method of electric heavy-duty truck motors.
[0032] Figure 2 This is a model predictive control principle diagram of the temperature and torque suppression control method for electric heavy-duty truck motors. Detailed Implementation
[0033] The technical solutions of the embodiments of the present invention will be described in further detail below with reference to the accompanying drawings. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention without creative effort are within the scope of protection of the present invention.
[0034] like Figures 1-2 As shown, an electric heavy-duty truck motor temperature and torque suppression control method utilizes on-board sensors to collect vehicle operating data in real time; the collected vehicle operating data includes, but is not limited to, ambient temperature T. env Vehicle load M veh Vehicle operating conditions S act (Uphill / Flat Road), Vehicle Acceleration V a Driving speed V b Coolant temperature T COOLANT Current I, voltage U.
[0035] By preprocessing the collected data to remove outliers and failure values, and combining data analysis algorithms, the motor temperature change trend is predicted in real time. A motor temperature prediction model is constructed based on historical vehicle operating data and neural network algorithms to predict future motor temperature changes according to the motor's operating status and external environmental conditions. An intelligent control strategy is designed using a recurrent neural network; when the predicted motor temperature exceeds a safe threshold, the motor output torque is automatically reduced to decrease the motor load and prevent overheating. By integrating the motor temperature prediction model with the motor torque control system, motor torque fluctuations are mitigated. Based on factors such as ambient temperature, vehicle operating conditions, coolant temperature, vehicle acceleration, speed, and gradient, the motor torque output is dynamically adjusted to maximize motor efficiency, achieving comprehensive optimization and automation of over-temperature torque limiting control for electric heavy-duty truck motors.
[0036] After real-time data acquisition and preprocessing, an SVM (Support Vector Machine) model is trained. A suitable kernel function (e.g., radial basis function kernel) and related parameters (such as the penalty factor C and kernel parameters) are selected, and parameter tuning is performed using methods such as cross-validation. During training, time-series data can be used to consider the characteristics of time series, such as the impact of previous temperature, load, and gradient on the current motor torque, to improve model accuracy. Based on the objective function and constraints, state variable estimation (x) is performed by optimizing the solver. k+1 =f(x) k ,u k ,d k The calculation of the optimal control quantity is then performed. Finally, the first step of the calculated control quantity is input into the system to achieve the motor torque threshold Z. T The optimized control motor temperature prediction model establishes a balance relationship between motor temperature and motor output torque. The balance relationship formula is described as follows:
[0037]
[0038] in:
[0039] L - A balance between meeting the motor temperature limit and maximizing the motor output torque;
[0040] T s,t -When the vehicle is operating under condition s, the predicted motor temperature at time t;
[0041] Z s,t -When the vehicle is operating under condition s, the maximum output torque of the motor within the motor temperature limit range at time t;
[0042] λ1, λ2 - Weighting coefficients for motor temperature and motor output torque.
[0043] The motor temperature prediction model incorporates ambient temperature T env Vehicle load M veh Vehicle operating conditions S act Vehicle acceleration V a Driving speed V b Coolant temperature T COOLANT The analysis models for current I and voltage U include, but are not limited to, model predictive control based on vector machines, recurrent neural networks, data-driven models, and physics-driven models. The analysis models are used at different ambient temperatures T. env and vehicle operating conditions S act This provides technical support for the motor torque control strategy.
[0044] Based on a multi-objective optimization model, a model predictive control algorithm is used to predict the future temperature of the motor. Under the condition of meeting the motor's operating temperature, intelligent optimal control of the motor's output torque is achieved. By dynamically adjusting the motor's output torque, the safe and reliable operation of the entire vehicle is ensured. The state variable x in the model predictive control of the motor temperature prediction model is... k Includes, but is not limited to, current I, voltage U, and control quantity u. k Including but not limited to motor temperature T MOT Disturbance d k Including but not limited to ambient temperature T env Vehicle load M veh Vehicle operating conditions S act Vehicle acceleration V a Driving speed V b Coolant temperature T COOLANT The output includes, but is not limited to, the motor torque threshold Z. T The intermediate quantity m of model predictive control k Including but not limited to motor operating temperature T mot Motor operating power P mot Motor speed N mot .
[0045] Model predictive control acquires data in real time, including the system's disturbance quantity d. k With state variable x k The collected data is input into the motor temperature prediction model for calculation. Based on the objective function and constraints, the state variables x are estimated through an optimization solver. k+1 =f(x) k ,u k ,d k ) and optimal control quantity u k The calculation is performed; finally, the calculated control quantity u is... k The first step is to input the motor torque control system, which is used for optimized control of the motor output torque.
[0046] The motor temperature prediction model sets constraint one during the model predictive control process: to keep the motor operating temperature within the over-temperature limit range, and the motor temperature T... MOT ≤ Maximum motor temperature T max This allows the motor torque control system to adjust the motor operating temperature (Tmot) to operate within a temperature range below the motor's limit temperature value, thereby preventing the motor from overheating and negatively impacting its lifespan and performance.
[0047] The motor temperature prediction model sets constraint condition two during the model predictive control process: maintaining the actual torque Z of the motor operation. Tmove ≥ Motor torque threshold Z T Battery load P bat ≤ Maximum output power of the battery P bat,max This ensures that the motor can operate reliably and safely when both the predicted operating temperature and the predicted torque threshold of the motor meet the operating requirements.
[0048] The objective function employs a multi-objective optimization model prediction objective function. The model predictive control strategy defines a suitable objective function minL, which typically includes the motor operating temperature T. mot and meet the motor operating temperature T mot The target is the optimal output torque of the motor, or a combination of both. The optimal output torque of the motor is determined by the ambient temperature T. env and vehicle operating conditions S act The decision was made that the motor output torque is maximized when the motor's operating temperature is met. Combining the temperature and torque targets satisfies the performance requirements of the electric heavy-duty truck motor and the safe and reliable operation of the entire vehicle. By solving this objective function, the system obtains the optimal motor output torque and implements adjustments accordingly.
[0049] Recurrent neural networks (RNNs) serve as the mathematical prediction model for model predictive control. The neural network receives the current system disturbance d. k State variable x k Control quantity u k The neural network takes ambient temperature T and other intermediate quantities as input and outputs a predicted future system state. The neural network design employs a multi-layer structure, with inputs including ambient temperature T. env Vehicle load M veh Vehicle operating conditions (climbing / flat road) S act Vehicle acceleration V a Driving speed V b Coolant temperature T COOLANT The system collects data such as current (I) and voltage (U), and outputs the optimal motor temperature. It then combines the processed collected data to train an SVM model, outputting the dynamically adjusted optimal motor torque threshold to achieve intelligent and high-precision control of the motor torque.
[0050] Motor operating temperature T mot With ambient temperature T env Vehicle load M veh Vehicle operating conditions S act The corresponding relationships are calibrated in advance, and the value of the motor output torque is calculated based on these relationships.
[0051] To train and optimize the neural network, it is necessary to collect system operation data of electric heavy-duty trucks under different operating conditions, including different ambient temperatures and thermal demands, including ambient temperature T. env Vehicle load M veh Vehicle operating conditions (climbing / flat road) S act Vehicle acceleration V a Driving speed V b Coolant temperature T COOLANT Current I, voltage U, and motor operating temperature T mot Motor operating power P mot Motor torque Z mot Motor speed N mot wait.
[0052] The model first needs to collect data, using the onboard system and other sensors to monitor ambient temperature, vehicle load, and road conditions (such as gradient) in real time. Then it enters the preprocessing stage. This involves cleaning the data, removing outliers, and applying filtering controls; extracting features related to motor performance (such as temperature change rate, load change rate, gradient, etc.); and normalizing the data to ensure consistent numerical ranges for different features, which is beneficial for model training.
[0053] The SVM model is trained using a preprocessed dataset. A suitable kernel function (e.g., radial basis function kernel) and related parameters (e.g., penalty factor C and kernel parameters) are selected, and parameter tuning is performed using methods such as cross-validation. During training, time-series data can be used to consider the effects of time-series characteristics, such as the temperature, load, and gradient of the previous moment on the current motor torque, to improve model accuracy. During vehicle operation, environmental data is collected in real time and input into the trained SVM model for prediction. The prediction result can be the expected output value of the motor torque. Based on the prediction result, the motor torque control strategy is adjusted to meet current needs and constraints. The control strategy can be an incremental adjustment based on fuzzy logic or PID control, converting the expected output value into an actual control signal and sending it to the motor control system.
[0054] The system monitors parameters such as motor temperature, battery load, and motor operating time in real time. Based on the monitoring results, it provides real-time feedback and adjusts control strategies accordingly. For example, if the motor temperature exceeds a set threshold, the motor torque output is reduced; if the battery load is too heavy, the vehicle load is reduced or the route is adjusted to reduce the gradient. Real-time feedback can be achieved through an onboard display or a remote monitoring system, allowing drivers or transport managers to understand the motor status at any time and take appropriate measures.
[0055] This invention combines historical vehicle operation data and data collected by an onboard real-time data monitoring system with a recurrent neural network motor temperature prediction model to output a predicted motor temperature. By combining an SVM model and a real-time monitoring system, dynamic adjustment of motor torque control is achieved, effectively avoiding excessive motor temperature and battery overload, while extending motor operating time and lifespan, and ensuring safe and reliable vehicle operation.
[0056] This invention is not limited to the above-described embodiments. Anyone should know that any structural changes made under the guidance of this invention, and any technical solutions that are the same as or similar to this invention, fall within the protection scope of this invention.
[0057] The technologies, shapes, and structures not described in detail in this invention are all known technologies.
Claims
1. A method for temperature and torque suppression control of an electric heavy-duty truck motor, characterized in that, Includes the following steps: S1: Real-time vehicle operation data is collected using onboard sensors; S2: Preprocess the collected data to remove outliers and failure values, and combine data analysis algorithms to predict the motor temperature change trend in real time; S3: Based on historical vehicle operation data and neural network algorithms, a motor temperature prediction model is constructed to predict future temperature changes of the motor according to the motor's operating status and external environmental conditions. S4: Combining recurrent neural networks to design intelligent control strategies, when the motor temperature is predicted to exceed the safety threshold, the motor output torque is automatically reduced; S5: By integrating the motor temperature prediction model with the motor torque control system, the motor torque output is dynamically adjusted to smooth out motor torque fluctuations.
2. The method for temperature and torque suppression control of electric heavy-duty truck motors according to claim 1, characterized in that, The vehicle operation data collected in step S1 includes, but is not limited to, ambient temperature T. env Vehicle load M veh Vehicle operating conditions S act Vehicle acceleration V a Driving speed V b Coolant temperature T COOLANT Current I, voltage U.
3. The method for temperature and torque suppression control of electric heavy-duty truck motors according to claim 2, characterized in that, The motor temperature prediction model in step S2 establishes a balance relationship between motor temperature and motor output torque. The balance relationship formula is described as follows: in: L - A balance between meeting the motor temperature limit and maximizing the motor output torque; T s,t -When the vehicle is operating under condition s, the predicted motor temperature at time t; Z s,t -When the vehicle is operating under condition s, the maximum output torque of the motor within the motor temperature limit range at time t; λ1, λ2 - Weighting coefficients for motor temperature and motor output torque.
4. The method for temperature and torque suppression control of electric heavy-duty truck motors according to claim 3, characterized in that, The motor temperature prediction model incorporates ambient temperature T. env Vehicle load M veh Vehicle operating conditions S act Vehicle acceleration V a Driving speed V b Coolant temperature T COOLANT The analysis models for current I and voltage U include, but are not limited to, model predictive control based on vector machines, recurrent neural networks, data-driven models, and physics-driven models. The analysis models are used at different ambient temperatures T. env and vehicle operating conditions S act This provides technical support for the motor torque control strategy.
5. The method for temperature and torque suppression control of electric heavy-duty truck motors according to claim 4, characterized in that, The state variable x in the motor temperature prediction model predicts control. k Includes, but is not limited to, current I, voltage U, and control quantity u. k Including but not limited to motor temperature T MOT Disturbance d k Including but not limited to ambient temperature T env Vehicle load M veh Vehicle operating conditions S act Vehicle acceleration V a Driving speed V b Coolant temperature T COOLANT The output includes, but is not limited to, the motor torque threshold Z. T The intermediate quantity m of model predictive control k Including but not limited to motor operating temperature T mot Motor operating power P mot Motor speed N mot ; Model predictive control acquires data in real time, including the system's disturbance quantity d. k With state variable x k The collected data is input into the motor temperature prediction model for calculation. Based on the objective function and constraints, the state variables x are estimated through an optimization solver. k+1 =f(x) k ,u k ,d k ) and optimal control quantity u k The calculation is performed; finally, the calculated control quantity u is... k The first step is to input the motor torque control system, which is used for optimized control of the motor output torque.
6. The method for temperature and torque suppression control of electric heavy-duty truck motors according to claim 5, characterized in that, The motor temperature prediction model sets constraint one during the model predictive control process: to keep the motor operating temperature within the over-temperature limit range, and the motor temperature T... MOT ≤ Maximum motor temperature T max This allows the motor torque control system to adjust the motor operating temperature Tmot according to actual needs, ensuring that the motor operates within a temperature range below the motor's limit temperature value.
7. The method for temperature and torque suppression control of electric heavy-duty truck motors according to claim 5, characterized in that, The motor temperature prediction model sets constraint condition two during the model predictive control process: maintaining the actual torque Z of the motor operation. Tmove ≥ Motor torque threshold Z T Battery load P bat ≤ Maximum output power of the battery P bat,max This ensures that the motor can operate reliably and safely when both the predicted operating temperature and the predicted torque threshold of the motor meet the operating requirements.
8. The method for temperature and torque suppression control of electric heavy-duty truck motors according to claim 5, characterized in that, The objective function is a multi-objective optimization model prediction objective function. The model predictive control strategy defines a suitable objective function minL, which typically includes the motor operating temperature T. mot and meet the motor operating temperature T mot The target is the optimal output torque of the motor, or a combination of both. The optimal output torque of the motor is determined by the ambient temperature T. env and vehicle operating conditions S act Decide.
9. The method for temperature and torque suppression control of an electric heavy-duty truck motor according to claim 5, characterized in that, The recurrent neural network in step S4 serves as the mathematical prediction model for model predictive control. The neural network receives the current system disturbance d. k State variable x k Control quantity u k Using other intermediate quantities as input, the output is a predicted future system state.
10. The method for temperature and torque suppression control of an electric heavy-duty truck motor according to claim 5, characterized in that, The motor operating temperature T mot With ambient temperature T env Vehicle load M veh Vehicle operating conditions S act The corresponding relationships are calibrated in advance, and the value of the motor output torque is calculated based on these relationships.
