A multi-timescale energy management method and system for a dual-rotor hybrid electric vehicle

Abstract

The present invention discloses a multi-timescale energy management method and system for a dual-rotor hybrid electric vehicle, belonging to the technical field of hybrid electric vehicle energy control. The method of the present invention first constructs a multi-timescale framework to optimize the vehicle's fuel economy on a long timescale and its dynamic performance on a short timescale. The long-timescale energy management strategy solves the nonlinear optimization problem using an improved alternating direction multiplier method, achieving optimal power distribution within the prediction range and improving computational efficiency. The short-timescale energy management strategy estimates the internal combustion engine output torque and observes the dual-rotor motor coupling torque, using an external motor to compensate for the additional power generated by the internal combustion engine speed regulation, thereby ensuring the stability of the output shaft torque.

Description

Technical Field

[0001] The present invention belongs to the technical field of hybrid electric vehicle energy control, and in particular relates to a multi-time-scale energy management method and system for a dual-rotor hybrid electric vehicle. Background Art

[0002] Dual-rotor hybrid systems have the potential to improve the efficiency, reliability, and compactness of powertrain operations. Compared to other dual-rotor motors, permanent magnet reluctance dual-rotor motors have better thermal and mechanical stability, high power density, and are suitable for applications that require frequent starting and disconnection. However, the complex electromagnetic coupling between the inner and outer motors increases electromagnetic losses and reduces output torque. In addition, the lower moment of inertia of the dual-rotor motor may cause changes in the operating point of the internal combustion engine. These transient characteristics not only affect the dynamic performance of the vehicle, but also affect the power allocation of the energy management strategy. Existing energy management strategies are usually based on steady-state conditions and are long-term optimizations, ignoring the above-mentioned transient response characteristics, which can lead to real-time energy imbalance.

[0003] To address these issues, two main technical directions are currently being explored: multi-objective optimization and integration with coordinated control. However, both approaches still present several challenges. Solving multi-objective optimization problems and the additional control degrees of freedom can increase the computational burden of energy management systems, complicating their real-time application. Furthermore, fuel economy and dynamic performance are optimization objectives at different timescales, resulting in a trade-off between the two optimization outcomes. Optimizing dynamic performance over long timescales can make it difficult to accurately simulate the transient response of the powertrain. Consequently, optimizing fuel economy over short timescales imposes additional computational burden. Multi-timescale frameworks can optimize the necessary combination of fast and slow system dynamics in a reasonable computational time while maintaining high accuracy, and have been widely used in energy management of integrated energy systems. In a dual-rotor hybrid system, the dynamic characteristics of the internal combustion engine and the dual-rotor motor differ significantly, and their dynamic characteristics cannot be accurately described over long timescales, particularly the electromagnetic coupling between the inner and outer motors in the dual-rotor motor. Furthermore, to achieve optimal power allocation within the prediction horizon, a long fixed-step optimization window must be used, which can result in long computational times. On the other hand, although the superiority of the alternating direction multiplier method embedded in the model predictive control framework has been demonstrated in fuel cell hybrid vehicles, parallel hybrid vehicles and power split hybrid vehicles, there are relatively few studies on the convex model of dual-rotor hybrid vehicles.

[0004] After searching the prior art, it was found that Chinese patent document No. CN114954422A, published on August 30, 2022, discloses a multi-time-scale multi-energy hybrid power system management and multi-controller coordination method, and proposes an upper-level energy management layer and a lower-level actuator control subsystem for a multi-energy hybrid power system. The energy management layer performs longitudinal control at different time scales based on the different dynamic characteristics of fuel cells, supercapacitors and batteries; at the same time, the upper-level energy management layer and the lower-level actuator subsystem perform lateral multi-time-scale control. Chinese patent document No. CN116154749A, published on May 23, 2023, discloses a hierarchical control method for a high-average-peak ratio hybrid energy system of a multi-electric aircraft, and proposes a hierarchical optimization method for the operation planning problem of a multi-electric aircraft. Different optimization objectives are optimized at different time scales, and then the optimization objectives are coordinated to achieve the optimal operation control of the multi-electric aircraft system. However, the above method is mainly aimed at the energy storage and power generation systems in the all-electric system. During the operation of the dual-rotor hybrid system, there is a problem of internal electromagnetic coupling of the dual-rotor motor, and its coupling situation is different under different operating conditions; in addition, while the internal combustion engine undertakes the task of power generation, it also transmits its output torque through the magnetic field of the internal motor. When the internal motor adjusts the speed of the internal combustion engine, it will also bring additional power requirements. These problems cannot be taken into account by the energy management layer. Summary of the Invention

[0005] The purpose of the present invention is to overcome the shortcomings of the above-mentioned prior art and provide a multi-time-scale energy management method and system for a dual-rotor hybrid vehicle to solve the shortcomings of the existing energy management methods of dual-rotor hybrid systems in the prior art, which cannot achieve coordinated optimization of fuel economy and dynamic performance and ignore the impact of dual-rotor electromagnetic coupling on energy management strategies.

[0006] In order to achieve the above object, the present invention adopts the following technical solutions:

[0007] A multi-timescale energy management method for a dual-rotor hybrid electric vehicle comprises the following steps:

[0008] S1, for dual-rotor hybrid vehicles, construct a steady-state dual-rotor motor power model and a transient dual-rotor motor torque model, and further obtain a plug-in hybrid vehicle system model of the dual-rotor motor;

[0009] S2, based on the current vehicle operating conditions, predicts the vehicle speed using a back-propagation neural network optimized by an improved particle swarm algorithm. Within the predicted speed range, the plug-in hybrid vehicle system model with a dual-rotor motor is convexified. The convexified plug-in hybrid vehicle system model is solved using an improved alternating direction multiplier method to obtain the optimal internal combustion engine power and optimal internal combustion engine operating curve, and further obtain the internal combustion engine set torque and dual-rotor motor set torque of the twin hybrid vehicle;

[0010] S3, constructing an internal combustion engine transient model and a dual-rotor motor coupling torque model for a plug-in hybrid vehicle system model with a dual-rotor motor; using the internal combustion engine transient model to monitor and estimate the state of the plug-in hybrid vehicle system in real time, taking the given torque of the internal combustion engine of the twin hybrid vehicle as the target, combined with the dual-rotor motor coupling torque model, the internal combustion engine speed and torque, as well as the dual-rotor motor torque are obtained.

[0011] A further improvement of the present invention is:

[0012] Preferably, in S1, the steady-state-based dual-rotor motor power model is:

[0013] (2)

[0014] Where, and They are the internal and external motor power, and are the average coupling torques of the internal and external motors in a long time scale, and are the outer rotor speed and the internal combustion engine speed, respectively. is the sign function, and are the internal and external motor efficiencies, respectively.

[0015] Preferably, in S1, the transient-based dual-rotor motor torque model is:

[0016] (1)

[0017] Where, is a table lookup function, is the torque of the dual-rotor motor given by the long-time-scale energy management strategy, and They are the inner rotor three-phase winding current and the stator current; and are the inner and outer rotor positions respectively.

[0018] Preferably, in S2, within the predicted vehicle speed range, the power balance relationship of the plug-in hybrid electric vehicle system model of the dual-rotor motor after convexification is:

[0019] (6)

[0020] Where, is the output power of the dual-rotor motor, Output power to the battery.

[0021] Preferably, in S2, the objective function of the plug-in hybrid electric vehicle system model with the convex dual-rotor motor is:

[0022] (11)

[0023] Where, is the prediction range, is the time step, is the battery output power, The battery charge state.

[0024] Preferably, in S2, the process of solving the optimized convexified plug-in hybrid electric vehicle system model of the dual-rotor motor by the improved alternating direction multiplier method is as follows: first, an inequality constraint is added to the objective function, the optimization objective function is reconstructed, the optimization objective function is Lagrangian expanded to obtain the Lagrangian function, the parameters and penalty factors in the Lagrangian function are updated by the alternating direction multiplier method, and the solution is completed when the set conditions or number of iterations are reached.

[0025] Preferably, the reconstruction optimization objective function is:

[0026] (11)

[0027] Where, is the prediction range, is the time step, is the battery output power, The battery charge state.

[0028] Preferably, in S2, the optimal torque and speed of the internal combustion engine are obtained based on the optimal internal combustion engine power and the optimal working curve of the internal combustion engine, and the optimal torque of the internal combustion engine is the given torque of the internal combustion engine; the given torque of the dual-rotor motor includes the given torque of the inner motor and the given torque of the outer motor, and the given torque of the inner motor is the optimal torque of the internal combustion engine, and the given torque of the outer motor is the difference between the required torque and the given torque of the inner motor.

[0029] Preferably, the specific process of S3 is: the internal combustion engine transient model collects the speed and electromagnetic speed of the inner motor in the dual-rotor motor in real time to estimate the output torque of the internal combustion engine, and uses the fast response characteristics of the outer motor to compensate for the output torque ripple and output torque lag of the internal combustion engine to obtain the internal combustion engine speed and torque; during the start-up of the internal combustion engine, the outer motor compensates the additional power to the inner motor.

[0030] A multi-time-scale energy management system for a dual-rotor hybrid electric vehicle, comprising:

[0031] The model building unit is used to build a steady-state dual-rotor motor power model and a transient-state dual-rotor motor torque model for a dual-rotor hybrid vehicle, and further obtain a plug-in hybrid vehicle system model of the dual-rotor motor.

[0032] The long-term unit is used to predict the vehicle speed based on the current vehicle operating conditions using a back-propagation neural network optimized by an improved particle swarm algorithm. Within the predicted speed range, the plug-in hybrid vehicle system model with a dual-rotor motor is convexified. The convexified plug-in hybrid vehicle system model is solved using an improved alternating direction multiplier method to obtain the optimal internal combustion engine power and optimal internal combustion engine operating curve, and further obtain the internal combustion engine set torque and dual-rotor motor set torque of the twin hybrid vehicle.

[0033] The transient unit is used to construct the internal combustion engine transient model and the dual-rotor motor coupling torque model of the plug-in hybrid vehicle system model with a dual-rotor motor; the internal combustion engine transient model is used to monitor and estimate the state of the plug-in hybrid vehicle system in real time. Taking the given torque of the internal combustion engine of the twin hybrid vehicle as the target, combined with the dual-rotor motor coupling torque model, the internal combustion engine speed and torque, as well as the dual-rotor motor torque are obtained.

[0034] Compared with the prior art, the present invention has the following beneficial effects:

[0035] This invention discloses a multi-timescale energy management method for a dual-rotor hybrid electric vehicle (HEV). It addresses the coordinated optimization of fuel economy and dynamic performance, as well as the real-time application of energy management strategies. The method first constructs a multi-timescale framework to optimize vehicle fuel economy on a long timescale and vehicle dynamic performance on a short timescale. The long-timescale energy management strategy solves the nonlinear optimization problem using an improved alternating direction multiplier method, achieving optimal power allocation within the prediction range and improving computational efficiency. The short-timescale energy management strategy estimates the internal combustion engine (ICE) output torque and observes the coupled torque of the dual-rotor motor. The external motor compensates for the additional power generated by the ICE speed regulation, ensuring stable output shaft torque. Simulation and experimental results demonstrate that, compared with other energy management methods, the proposed method optimizes vehicle performance on two timescales, improving dynamic performance while maintaining fuel economy and ensuring real-time application. By integrating different optimization objectives on two timescales and fully considering the electromagnetic coupling of the dual-rotor motor and the dynamic characteristics of the ICE, the proposed method achieves coordinated optimization of system fuel economy and dynamic performance, with the potential for real-time application.

[0036] Furthermore, on a long time scale, the present invention significantly improves the fuel economy of the vehicle by using a nonlinear model predictive control energy management optimized by an improved alternating direction multiplier method. Compared with the rule-based energy management strategy, fuel consumption is reduced by 25.05% and the fuel consumption rate is reduced by 15.88%. On a short time scale, the output shaft torque is optimized through composite torque coordination, and the output shaft torque fluctuation is reduced by 45.24%. BRIEF DESCRIPTION OF THE DRAWINGS

[0037] Figure 1 Schematic diagram of the dual-rotor hybrid system structure.

[0038] Figure 2 Schematic diagram of the multi-time-scale energy management strategy of a dual-rotor hybrid system.

[0039] Figure 3 Illustration of coordinated control of control sequences at different time scales.

[0040] Figure 4 Schematic diagram of the short-time-scale composite torque coordination control strategy.

[0041] Figure 5 Comparison chart of internal combustion engine operating points under different optimization algorithms.

[0042] Figure 6 This is a comparison chart of dynamic performance optimization effects. DETAILED DESCRIPTION

[0043] Hereinafter, the terms "first," "second," "third," and "fourth" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the quantity of the technical features indicated. Thus, a feature identified as "first," "second," "third," or "fourth" may explicitly or implicitly include one or more of such features.

[0044] The co-shooting method provided in the embodiments of the present application can be applied to terminal devices such as mobile phones, tablet computers, wearable devices, vehicle-mounted devices, augmented reality (AR) / virtual reality (VR) devices, laptop computers, ultra-mobile personal computers (UMPCs), netbooks, and personal digital assistants (PDAs). The embodiments of the present application do not impose any restrictions on the specific types of terminal devices.

[0045] It should be noted that the terms "first," "second," and the like in the description and drawings of the present invention are used to distinguish similar objects and are not necessarily used to describe a specific order or precedence. It should be understood that the terms used in this manner are interchangeable where appropriate, so that the embodiments of the present invention described herein can be implemented in an order other than those illustrated or described herein. In addition, the terms "including" and "having," as well as any variations thereof, are intended to cover non-exclusive inclusions. For example, a process, method, system, product, or apparatus comprising a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units that are not explicitly listed or that are inherent to these processes, methods, products, or apparatus.

[0046] This invention provides an energy management method for a dual-rotor hybrid powertrain system, optimizing both vehicle fuel economy and dynamic performance across multiple timescales. On a long-term scale, the method uses current vehicle operating information to predict future vehicle speeds. The optimization objective function is then solved within the predicted range to achieve optimal power distribution within that range. On a short-term scale, a transient response model for the vehicle powertrain is constructed, including an internal combustion engine torque estimate and a model for the transient coupling torque of the dual-rotor motor. This method achieves stable output shaft torque by compensating for engine torque loss and coupling torque.

[0047] The specific steps include:

[0048] S1: Based on the structural characteristics of the dual-rotor hybrid system, the impact of electromagnetic coupling between the internal and external motors of the dual-rotor motor on energy management and system dynamic performance is analyzed. A plug-in dual-rotor hybrid vehicle model and a permanent magnet reluctance dual-rotor motor electromagnetic coupling model are constructed. The constructed system model will be used as a predictive model in model predictive control on a long time scale and as a coupling torque observation on a short time scale. The specific process is as follows:

[0049] Finite element simulations were used to analyze the impact of electromagnetic coupling within a permanent magnet reluctance dual-rotor motor on its average output torque and efficiency under different operating conditions. A transient torque model and a steady-state power model for the dual-rotor motor were constructed, accounting for electromagnetic coupling. Then, based on the structural characteristics of the dual-rotor hybrid system, a plug-in hybrid vehicle model based on the permanent magnet reluctance dual-rotor motor was constructed, referred to as the dual-rotor hybrid system model in the following description. The dual-rotor hybrid system model primarily includes an internal combustion engine model, a dual-rotor motor model, a battery model, a vehicle dynamics model, and a transmission system model. The internal combustion engine model uses a traditional lookup table model, the battery model is a series structure of an ideal voltage source and a variable resistor, the vehicle longitudinal dynamics model is used for the vehicle dynamics model, and the transmission system model is a mechanism model. Unlike traditional dual-rotor motor modeling, the dual-rotor motor model established in this paper, which accounts for coupling, analyzes the impact of electromagnetic coupling within the dual-rotor motor on the system's output torque and efficiency. The motor power model of the traditional dual-rotor motor model was reconstructed, improving the model's accuracy.

[0050] Among them, the transient torque model of the dual-rotor motor considering electromagnetic coupling is shown in formula (1):

[0051] (1)

[0052] Where, is a table lookup function, is the torque of the dual-rotor motor given by the long-time-scale energy management strategy, and They are the inner rotor three-phase winding current and the stator current; and are the inner and outer rotor positions respectively.

[0053] The steady-state power model of the dual-rotor motor considering electromagnetic coupling is shown in formula (2):

[0054] (2)

[0055] Where, and They are the internal and external motor power, and are the average coupling torques of the internal and external motors in a long time scale, and are the outer rotor speed and the internal combustion engine speed, respectively. is the sign function, and are the internal and external motor efficiencies, respectively. Due to the electromagnetic coupling between the internal and external motors, the internal motor efficiency is not only related to its operating point, but also affected by the external motor's operating state; the same is true for the external motor. Therefore, the internal and external motor efficiencies are shown in formula (3):

[0056] (3)

[0057] Where, and They are the lookup table functions for the internal and external motor efficiencies respectively.

[0058] In the constructed S2 and S3, a multi-time-scale energy management system was constructed, e.g. Figure 2 As shown, it includes a long-time scale energy management system with fuel economy as the optimization target and a short-time scale energy management system with dynamic performance as the optimization target.

[0059] Coordination of control sequences at long and short time scales Figure 3 At the current moment t On a long-time scale, the optimal power allocation within the prediction range is first determined based on the system operating state. This first element is used as the system input and the given input for short-time-scale energy management. Then, by modeling and monitoring the system's transient response on a short-time scale, the system's real-time output torque is compensated to ensure a stable output torque.

[0060] S2. In the long-time-scale energy management system, a nonlinear model predictive control energy management system based on the improved alternating direction multiplier method is constructed, and rolling optimization is performed in combination with the optimization objectives, system constraints and system status to obtain the optimal power distribution of the system.

[0061] S201: Based on the current vehicle operation information, the back propagation neural network optimized by the improved particle swarm algorithm is used to predict the future driving conditions of the vehicle, and the required driving power and torque are calculated based on the vehicle longitudinal dynamics model. The speed prediction model is shown in formula (4):

[0062] (4)

[0063] Where, For the nonlinear relationship of the prediction model, The future i The predicted speed at each moment, is the forecast range.

[0064] S202: Within the speed prediction range, the dual-rotor hybrid power system model is convexified, and the optimization target and system constraints of the dual-rotor hybrid power system model are reconstructed.

[0065] The optimization objective of a dual-rotor hybrid system is non-convex, making the real-time application of energy management algorithms dependent on the speed of the optimization problem solution. To reduce the solution time and improve the fuel economy of the hybrid system, the nonlinear model of the dual-rotor hybrid system is convexified and the optimization problem is solved using an improved alternating direction multiplier method.

[0066] Considering the influence of electromagnetic coupling, the power model of the dual-rotor motor can be simulated by a quadratic function between the input electrical power and the output mechanical power. The input power of the inner motor and the input power of the outer motor are shown in formula (5).

[0067] (5)

[0068] Where, and are the inner and outer motor speeds respectively, and are the inner and outer motor torques respectively, and Output mechanical power to the inner and outer motors; and are the fitting coefficients, and , , and This is satisfied within the operating speed range of the dual-rotor motor.

[0069] At each time step within the predicted vehicle speed range, the output shaft speed and required drive power are known. Assuming the internal combustion engine operates at its optimal operating curve, the input-output relationship in Equation (5) can be reduced to a time-varying convex function of the output power.

[0070] Therefore, at the kth time step, the power balance relationship of the dual-rotor hybrid system is shown in formula (6):

[0071] (6)

[0072] Where, is the output power of the dual-rotor motor, is the battery output power, and the input-output relationship of the dual-rotor motor is shown in formula (7)

[0073] (7)

[0074] Where, is the fitting coefficient.

[0075] Assuming that the internal combustion engine operates at its optimal operating curve, the fuel consumption rate can be determined solely by the engine power The convex function representation of is shown in formula (8):

[0076] (8)

[0077] Where, are the fitting coefficients and , .

[0078] Assuming that the battery open circuit voltage and battery internal resistance do not change with the battery SOC, the battery dynamics are as shown in formula (9):

[0079] (9)

[0080] in, Output power to the battery; is the open circuit voltage of the battery; is the internal resistance of the battery.

[0081] In order to ensure that the above relationship is monotonically increasing, the system constraints are shown in formula (10)

[0082] (10)

[0083] In the formula, the subscript and Represent the lower and upper limits of the physical quantity constraints respectively.

[0084] Therefore, the reconstructed objective function is a convex function and is determined only by the battery power, as shown in formula (11),

[0085] (11)

[0086] Where, is the prediction range, is the time step, is the battery output power, The battery charge state.

[0087] S203: Using an improved alternating direction multiplier method to solve the convex optimization problem at each prediction step, the optimal internal combustion engine output power within the prediction range is obtained.

[0088] when When , there is enough energy in the battery at each time step within the prediction range to allow the motor to run at maximum power, then the solution of formula (11) is Otherwise, it is necessary to solve the optimization problem in formula (11). In order to improve the performance of real-time execution, the present invention proposes a nonlinear model predictive control based on the improved alternating direction multiplier method. At the same time, the virtual variable , by separating the variables to simplify the iteration of the solver. By adding inequality constraints to the objective function, formula (11) can be reconstructed into a new optimization objective function, as shown in formula (12),

[0089] (12)

[0090] In the formula, the indicator function The expression of is shown in formula (13),

[0091] (13)

[0092] The extended Lagrangian function of the optimization objective function is shown in formula (14),

[0093] (14)

[0094] Where, is the penalty factor, is the Lagrange multiplier, is a unit vector, For the element The lower triangular matrix of .

[0095] The parameter update in the alternating direction multiplier method is shown in formula (15):

[0096] (15)

[0097] In the formula

[0098] (16)

[0099] Penalty Factor The update rule of is shown in formula (17),

[0100] (17)

[0101] in, For the element The lower triangular matrix of is a dummy variable, is a unit vector.

[0102] Typically, the iterations of the alternating direction multiplier method are and , or stop when the number of iterations exceeds a set threshold. Error Usually determined in advance by the dimensionality of the variable.

[0103] S204: Based on the optimal internal combustion engine power and the dual-rotor hybrid system structure obtained in S203, the internal combustion engine is set to operate at an optimal operating curve to obtain the optimal system torque distribution, including the internal combustion engine set torque and the dual-rotor motor set torque. The specific process is as follows: Based on the internal combustion engine optimal operating curve and the internal combustion engine optimal power obtained in S203, the optimal internal combustion engine torque and speed are obtained; the internal motor set torque is equal to the optimal internal combustion engine torque, and the external motor set torque is the difference between the required torque and the internal motor set torque.

[0104] S3: If Figure 4 As shown, in the short-time-scale energy management system, considering the different dynamic response characteristics of the power system components and the electromagnetic coupling characteristics of the dual-rotor motor, a composite torque coordination control strategy is designed to coordinate the control system output torque. Combined with the power distribution results of the long-time-scale energy management system, the final hybrid system input reference torque and speed are obtained, specifically including the reference torque and speed of the internal combustion engine and the reference torque of the dual-rotor motor.

[0105] S301: Construct a system transient response model, and construct an internal combustion engine transient model and a dual-rotor motor coupling torque model on a short time scale.

[0106] S302: Monitor and estimate the hybrid system state in real time based on the system transient model, and perform torque coordination control under different operating conditions, including: internal combustion engine starting, internal combustion engine speed regulation, and hybrid drive mode. Specifically:

[0107] Based on the speed and electromagnetic torque of the inner motor in the dual-rotor motor, the output torque of the internal combustion engine is estimated, and the fast response characteristics of the outer motor are used to compensate for the output torque ripple and output torque lag of the internal combustion engine. According to the dual-rotor motor coupling torque model, the coupling torque is obtained by monitoring the inner and outer motor currents and the inner and outer rotor positions, and then compensated. During the internal combustion engine starting and speed regulation process, the inner motor requires additional power for regulation, and the additional power is compensated by the outer motor to ensure the stability of the output torque, as shown in formula (18).

[0108] (18)

[0109] Where, and For the given external motor torque and given internal combustion engine torque of the long-time energy management strategy, Estimate the torque for the internal combustion engine, For a given speed of the internal combustion engine, and The coupling compensation torque for the inner and outer motors.

[0110] After simulation and experimental analysis, the fuel economy comparison of the multi-time scale energy management, dynamic programming, nonlinear model predictive control and rule-based energy management strategies proposed in the present invention is shown in Table 1. Among them, the dynamic programming has the lowest fuel consumption, which is because the complete working condition information is known in advance. In addition, the effective fuel consumption rate of the proposed strategy is the lowest, because its working points are mostly concentrated in the low fuel consumption area. Compared with the nonlinear model prediction, the proposed multi-time scale energy management strategy modifies the working point of the internal combustion engine by considering the dynamic characteristics of the powertrain and electromagnetic coupling, so the internal combustion engine has higher working efficiency and lower fuel consumption. In addition, its fuel consumption is 95.35% similar to the offline theoretical optimal value, which is 25.05% lower than the rule-based energy management strategy. The operation of the system under different control strategies is shown in the figure. Figure 5 As shown in the figure, due to the battery capacity limitation, different methods all maintain the remaining battery capacity within a reasonable range. Compared with the theoretical optimal value obtained by dynamic programming, the proposed multi-time-scale energy management strategy is more inclined to maintain the battery capacity near its lower limit due to the limitation of its prediction range. In addition, compared with rule-based methods and nonlinear model predictive control, dynamic programming and the proposed strategy maintain the internal combustion engine power and motor torque within a wider range, thereby increasing the frequency of occurrence in the high-efficiency zone.

[0111] Dynamic performance optimization results are as follows Figure 6 As shown. Compared with the single-time-scale energy management strategy, the output shaft torque of the proposed method can quickly reach a stable value due to the coordinated control of the internal combustion engine speed regulation and power output. Since the expected power of the internal combustion engine suddenly increases, the internal combustion engine will not output torque at first, and the speed of the internal combustion locomotive can quickly reach the expected value through the internal motor. At the same time, the external motor quickly generates torque to compensate for the additional torque generated by the internal combustion engine speed regulation, which also ensures the stability of the internal combustion engine speed. On the other hand, considering the influence of electromagnetic coupling, the proposed strategy speeds up the response time of the external motor, thereby ensuring the stability of the output shaft torque. The proposed method can stabilize the output power of the internal combustion engine and reduce the output shaft torque fluctuation by 45.24%, thereby improving the dynamic performance of the vehicle.

[0112] Table 1 Fuel consumption under different control strategies

[0113]

[0114] In summary, the present invention relates to a multi-timescale energy management method for a dual-rotor hybrid system considering electromagnetic coupling. The method comprises a long-timescale nonlinear model predictive control energy management strategy based on an improved alternating direction multiplier method and a short-timescale composite torque coordinated control strategy. The long-timescale energy management strategy monitors the vehicle's operating status in real time, predicts the vehicle speed within a certain range in the future using a back-propagation neural network optimized by an improved particle swarm algorithm, collects the current vehicle speed and battery state of charge into the long-timescale energy management system, and solves the optimal power allocation within the constraints of the system's physical constraints, taking fuel economy as the optimization objective. To improve fuel economy and real-time application capabilities, the dual-rotor hybrid system is convexified and an improved alternating direction multiplier method is proposed to solve the optimization problem, significantly improving computational efficiency while ensuring fuel economy. The short-timescale energy management strategy establishes a system transient response model, collects the internal and external motor speeds, currents, and rotor positions in real time, estimates the internal combustion engine output torque, and compensates for its output torque ripple and output torque lag. Furthermore, when the internal motor regulates the internal combustion engine speed, the excess power results in a loss of output shaft torque. This reverse torque is compensated by the external motor to ensure smooth output torque. When the internal and external motors operate simultaneously, electromagnetic coupling causes the output torque of the internal and external motors to be less than the set torque. This coupling torque is observed and compensated using a transient coupling torque model. This composite torque coordination control strategy ensures smooth output shaft torque and improves vehicle dynamic performance.

[0115] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the scope of protection of the present invention.

Claims

1. A multi-time-scale energy management method for a dual-rotor hybrid electric vehicle, characterized in that: The following steps are involved: S1, for dual-rotor hybrid vehicles, construct a steady-state dual-rotor motor power model and a transient dual-rotor motor torque model, and further obtain a plug-in hybrid vehicle system model of the dual-rotor motor; S2, based on the current vehicle operating conditions, predicts the vehicle speed using a back-propagation neural network optimized by an improved particle swarm algorithm. Within the predicted speed range, the plug-in hybrid vehicle system model with a dual-rotor motor is convexified. The convexified plug-in hybrid vehicle system model is solved using an improved alternating direction multiplier method to obtain the optimal internal combustion engine power and optimal internal combustion engine operating curve, and further obtain the internal combustion engine set torque and dual-rotor motor set torque of the twin hybrid vehicle; S3, constructing an internal combustion engine transient model and a dual-rotor motor coupling torque model for a plug-in hybrid electric vehicle system model with a dual-rotor motor; using the internal combustion engine transient model to monitor and estimate the state of the plug-in hybrid electric vehicle system in real time, taking the given torque of the internal combustion engine of the dual-rotor hybrid electric vehicle as the target, combined with the dual-rotor motor coupling torque model, the internal combustion engine speed and torque, as well as the dual-rotor motor torque are obtained.

2. The multi-time-scale energy management method for a dual-rotor hybrid electric vehicle according to claim 1, characterized in that: In S1, the steady-state dual-rotor motor power model is: （2） Where, and They are the internal and external motor power, and are the average coupling torques of the internal and external motors in a long time scale, and are the outer rotor speed and the internal combustion engine speed, respectively. is the sign function, and are the internal and external motor efficiencies, respectively.

3. The multi-time-scale energy management method for a dual-rotor hybrid electric vehicle according to claim 1, characterized in that: In S1, the transient-based dual-rotor motor torque model is: （1） Where, is a table lookup function, is the torque of the dual-rotor motor given by the long-time-scale energy management strategy, and They are the inner rotor three-phase winding current and the stator current; and are the inner and outer rotor positions respectively.

4. The multi-time-scale energy management method for a dual-rotor hybrid electric vehicle according to claim 1, characterized in that: In S2, within the predicted vehicle speed range, the power balance relationship of the plug-in hybrid electric vehicle system model with a dual-rotor motor after convexification is: （6） Where, is the output power of the dual-rotor motor, Output power to the battery.

5. The multi-time-scale energy management method for a dual-rotor hybrid electric vehicle according to claim 1, characterized in that: In S2, the objective function of the plug-in hybrid electric vehicle system model with the convex dual-rotor motor is: （11） Where, is the prediction range, is the time step, is the battery output power, The battery charge state.

6. The multi-time-scale energy management method for a dual-rotor hybrid electric vehicle according to claim 5, characterized in that: In S2, the process of solving the optimized convexified plug-in hybrid electric vehicle system model of the dual-rotor motor by the improved alternating direction multiplier method is as follows: first, an inequality constraint is added to the objective function, the optimization objective function is reconstructed, the optimization objective function is Lagrangian expanded to obtain the Lagrangian function, the parameters and penalty factors in the Lagrangian function are updated by the alternating direction multiplier method, and the solution is completed when the set conditions or number of iterations are met.

7. The multi-time-scale energy management method for a dual-rotor hybrid electric vehicle according to claim 6, characterized in that: The reconstructed optimization objective function is: （11） Where, is the prediction range, is the time step, is the battery output power, The battery charge state.

8. The multi-time-scale energy management method for a dual-rotor hybrid electric vehicle according to claim 7, characterized in that: In S2, the optimal torque and speed of the internal combustion engine are obtained based on the optimal internal combustion engine power and the optimal working curve of the internal combustion engine. The optimal torque of the internal combustion engine is the given torque of the internal combustion engine; the given torque of the dual-rotor motor includes the given torque of the inner motor and the given torque of the outer motor. The given torque of the inner motor is the optimal torque of the internal combustion engine, and the given torque of the outer motor is the difference between the required torque and the given torque of the inner motor.

9. The multi-time-scale energy management method for a dual-rotor hybrid electric vehicle according to claim 1, characterized in that: The specific process of S3 is as follows: the internal combustion engine transient model collects the speed and electromagnetic speed of the inner motor in the dual-rotor motor in real time to estimate the output torque of the internal combustion engine, and uses the fast response characteristics of the outer motor to compensate for the output torque ripple and output torque lag of the internal combustion engine to obtain the internal combustion engine speed and torque; during the startup of the internal combustion engine, the outer motor compensates the additional power to the inner motor.

10. A multi-time-scale energy management system for a dual-rotor hybrid electric vehicle, characterized in that: include: A model building unit is used to build a steady-state dual-rotor motor power model and a transient-state dual-rotor motor torque model for a dual-rotor hybrid vehicle, and further obtain a plug-in hybrid vehicle system model of the dual-rotor motor; The long-term unit is used to predict the vehicle speed based on the current vehicle operating conditions using a back-propagation neural network optimized by an improved particle swarm algorithm. Within the predicted speed range, the plug-in hybrid vehicle system model with a dual-rotor motor is convexified. The convexified plug-in hybrid vehicle system model is solved using an improved alternating direction multiplier method to obtain the optimal internal combustion engine power and optimal internal combustion engine operating curve, and further obtain the internal combustion engine set torque and dual-rotor motor set torque of the twin hybrid vehicle. The transient unit is used to construct the internal combustion engine transient model and the dual-rotor motor coupling torque model of the plug-in hybrid vehicle system model with a dual-rotor motor; the state of the plug-in hybrid vehicle system is monitored and estimated in real time through the internal combustion engine transient model. Taking the given torque of the internal combustion engine of the dual-rotor hybrid vehicle as the target, the internal combustion engine speed and torque, as well as the dual-rotor motor torque are obtained in combination with the dual-rotor motor coupling torque model.

Images