Self-adaptive control method for motor-driven torsional vibration of planetary hybrid vehicle

Abstract

The invention provides a self-adaptive control method for motor driving torsional vibration of a planetary hybrid power vehicle, and belongs to an automobile control system, aiming at solving the problem that the torsional vibration in the motor driving process of the existing planetary multi-gear hybrid power vehicle is difficult to quickly reduce and the like. According to the method, torque reconstruction and feedback control are carried out on a motor torque demand obtained by an energy management strategy, and the self-adaptive adjusting torque is obtained together with optimal control parameters output by a constructed working condition control parameter library obtained by off-line optimization of control parameters; and after interference adaptive compensation is carried out, the total target torque of the motor is obtained and output to the controlled hybrid power vehicle. According to the method provided by the invention, control parameters can be automatically adjusted according to working condition changes, the motor control effect is improved, and meanwhile, an interference adaptive compensation algorithm is utilized, so that the anti-interference capability of a torque reconstruction and feedback control algorithm is improved to a certain extent.

Description

Technical Field

[0001] The present invention belongs to automobile control systems, and more specifically, to torsional vibration control during motor driving of planetary hybrid vehicle. Background Art

[0002] Planetary hybrid power systems have good power and economy and are widely used in the field of domestic commercial vehicles. In the case of planetary multi-speed hybrid power systems, the engine, dual motors and wet clutch are power coupled to the gearbox through a planetary gear mechanism. Due to the various states of the clutch transmission elements, when the output torque changes transiently, it is easy to cause abnormal problems of torque vibration in the transmission system. In addition, when the engine excitation frequency is close to the system's natural frequency, resonance may be triggered, which will have an adverse effect on the stability of the vehicle. In severe cases, it may even cause shaft breakage and damage to the torsional vibration damper. Therefore, it is necessary to urgently solve the problem of torque vibration in planetary multi-speed hybrid power systems. Aiming at the problem of transient torsional vibration of planetary hybrid vehicles, the comprehensive performance of the system, such as power, economy, comfort and safety, can be improved, which is conducive to promoting the successful application and promotion of planetary multi-speed hybrid products.

[0003] In recent years, in order to solve the torsional vibration problem of vehicle transmission system to meet the needs of practical engineering, domestic scholars have also carried out relevant research on torsional vibration adaptive control methods. For example, the Chinese patent publication number is CN116738809A, the publication date is 2013-09-12, and discloses a simulation modeling and parameter optimization method for the torsional vibration characteristics of multi-power source vehicle transmission system. The method uses multiple algorithms to optimize the parameters, obtains a high-precision kinematic simulation model of the vehicle transmission system, and realizes the optimization of transmission system vibration reduction. The Chinese patent publication number is CN111591144A, the publication date is 2020-08-28, and discloses a control method for reducing the output torque vibration of the motor for electric vehicles. The invention accurately judges the actual working conditions of the vehicle according to the corresponding vehicle and motor input signals and constructs the corresponding algorithm to select the appropriate filter coefficient and torque compensation coefficient by looking up the table to achieve the requirements of smoother and safer driving of the whole vehicle. However, the existing patents do not propose a precise method for controlling the torsional vibration of the motor drive of the planetary hybrid system under multiple working conditions. According to the characteristics of planetary hybrid power system, the corresponding motor-driven torsional vibration adaptive control method is formulated, which is of great significance to the improvement of the comprehensive performance of planetary hybrid vehicles, such as power, economy, comfort and safety. Summary of the invention

[0004] The technical problem to be solved by the present invention is to overcome the problem that the torsional vibration in the motor driving process of the existing planetary multi-speed hybrid vehicle is difficult to quickly reduce, and a method for adaptive control of torsional vibration of the motor driving of a planetary hybrid vehicle is provided.

[0005] In order to solve the above technical problems, the present invention is implemented by adopting the following technical solution: comprising the following steps:

[0006] The first step is torque reconstruction and feedback control

[0007] Firstly, the rate of change of the initial motor torque demand is corrected through the first-order inertia link to ensure that the oscillation amplitude of the longitudinal acceleration is reduced without affecting the vehicle's dynamics, and the motor reference torque is smoothed; then the motor reference torque is divided into three parts according to the ratio of the splitting coefficient, and two of the parts are delayed. The three partial torques can be superimposed on the principle of reducing system vibration by superposing non-cophase signals to obtain the reconstructed torque of the motor; the longitudinal acceleration calculated by the nominal model is used as the feedback quantity, and the difference between the real-time longitudinal acceleration and the expected longitudinal acceleration is calculated. When the real-time longitudinal acceleration reaches the expected total acceleration, the feedback torque is output.

[0008] The second step is to build the working condition control parameter library offline

[0009] Combined with the actual vehicle road test data, the torque change time and change difference are divided into 4*4 16 working conditions at equal intervals, covering the torque change range of the vehicle during actual driving; the particle swarm intelligent algorithm is used to reconstruct the torque and optimize the feedback control parameters offline. The particle value is the control parameter, which mainly includes the split coefficient, delay factor, and feedback gain; the current particle control parameters are input into the transmission system model to calculate the response, and the normalized maximum amplitude of vehicle acceleration and the maximum amplitude of the motor-wheel equivalent speed difference are used as the fitness function of the particle swarm algorithm. obj The optimization goal

[0010]

[0011] In the formula, X is the control parameter

[0012] a diff (X)——Maximum amplitude of longitudinal acceleration

[0013] ω diff (X)——Maximum amplitude of motor-wheel equivalent speed difference

[0014] a diff_ori ——The maximum amplitude of longitudinal acceleration corresponding to no control

[0015] ω diff_ori ——The maximum amplitude of longitudinal acceleration corresponding to no control The maximum amplitude of the motor-wheel equivalent speed difference

[0016] The individual optimal position and the global optimal position are recorded according to the size of the fitness; if the number of iterations is lower than the set maximum number of iterations and the fitness value is greater than the minimum fitness threshold, the speed and position of each particle are updated; when the number of iterations or the fitness value is less than the minimum fitness threshold, that is, when the termination condition is reached, the particle swarm optimization process ends and the optimal control parameter combination is output; the mapping relationship between the torque demand and the optimal control parameters is used to construct a working condition control parameter library.

[0017] The third step is to adaptively update the control parameters online.

[0018] The randomly input motor torque demand is used to identify the torque location point and calculate the weighted parameters, which are then loaded into the control parameter library for online calling of the optimal control parameters.

[0019] Step 4: Interference adaptive compensation

[0020] The Inverse Nominal Model Observer (INMO) algorithm based on the nominal inverse model is used for adaptive disturbance compensation. The transmission shaft speed fed back by the controlled vehicle is used as input. The disturbance is observed based on the physical implementation and filter design of the nominal inverse model, and the disturbance compensation amount is calculated to reduce the influence of uncertainty disturbance on the torsional vibration control effect. The six-degree-of-freedom torsional vibration dynamics model is simplified to derive the nominal model transfer function matrix, and the transmission system is equivalent to two inertia elements of the ring gear-reducer-differential and the whole vehicle. The two inertia elements are connected by an elastic element with equivalent stiffness and equivalent damping. The dynamic model of the transmission system is:

[0021]

[0022] Where, T AMT_out ——Gearbox output torque

[0023] T 11 ——Transmitted torque between two inertia elements

[0024] i0——speed ratio of main reducer

[0025] I 11 ——Equivalent moment of inertia of ring gear-reducer-differential inertia components

[0026] I 12 ——Equivalent moment of inertia of the vehicle’s inertial components

[0027] ——The angle of equivalent moment of inertia of ring gear-reducer-differential

[0028] ——The angle of equivalent moment of inertia of the vehicle

[0029] c 11 ——Equivalent damping of the elastic element between two inertial elements

[0030] k 11 ——Equivalent stiffness of the elastic element between the two inertial elements

[0031] T L1 ——The sum of driving resistance torque and braking torque

[0032] Write the model in standard form of state-space equations

[0033]

[0034] By performing Lapalus transformation on the state space equation, the transfer function matrix of the system can be calculated as

[0035] P(s)=C[sI-A] -1 B+D (4)

[0036] The transfer function of the nominal model can be expressed as

[0037]

[0038] According to the Mason gain formula, the transfer function from each input to output can be calculated as

[0039]

[0040] Where P a (s)——Adaptive torque transfer function

[0041] P b (s)——disturbance transfer function

[0042] P c (s)——Noise transfer function

[0043] P act (s)——transfer function of the true model

[0044] After obtaining the transfer function of the nominal model, it is necessary to introduce a filter after the interference estimation. The filter design principle can be summarized as follows: Q(s)≈1 should be achieved in the low-frequency range, and Q(s)≈0 should be achieved in the high-frequency range.

[0045] In the third step of the present invention, location point identification refers to and weighted calculation refers to:

[0046] By comparing the set 16 working conditions with the randomly input motor torque change time and torque change difference, 9 areas are formed; any input torque demand will fall into the 9 areas, and then the torque change time ratio and torque change difference ratio can be directly calculated to obtain the control output u adapt .

[0047]

[0048] In the formula, K req ——Weighting factor

[0049] Δt1, Δt2, ΔT req1 , ΔT req2 ——Set the torque change time and difference required by the working condition

[0050] Δt, ΔT req ——Input at any time

[0051] u 11 、u 21 、u 12 、u 22 ——The torque time variation values ​​corresponding to the lower left, lower right, upper left and upper right of the current area are compared with the prior art. The beneficial effects of the present invention are:

[0052] 1. The method for adaptive torsional vibration control of a motor drive of a planetary hybrid vehicle described in the present invention is to design a set of motor torque adaptive control functions based on the energy management strategy of a planetary hybrid system, without requiring excessive modification of the original system, and giving full play to the advantages of the planetary hybrid system in terms of economy and power.

[0053] 2. The method for adaptive torsional vibration control of a motor drive of a planetary hybrid vehicle described in the present invention uses a proportional splitting and partial delay method to reconstruct the motor torque, and uses the principle of superimposing non-co-phase signals to reduce system vibration to reconstruct the torque, thereby effectively reducing the vibration of the motor output torque;

[0054] 3. The method for adaptive control of torsional vibration of motor drive of a planetary hybrid vehicle of the present invention adopts a method for adaptive adjustment of control parameters under multiple working conditions, which can adaptively update vehicle control parameters and well adapt to the randomness of motor torque caused by different driver operations during actual driving of the vehicle;

[0055] 4. The planetary hybrid vehicle motor drive torsional vibration adaptive control method of the present invention adopts random input torque demand location point identification and weighted calculation, which can well identify the working condition characteristics of the current vehicle and ensure the robustness and accuracy of the control algorithm;

[0056] 5. The planetary hybrid vehicle motor drive torsional vibration adaptive control method described in the present invention adopts an interference adaptive compensation method, observes the interference based on the physical realization and filter design of the nominal inverse model, and calculates the interference compensation amount to reduce the influence of external interference on the algorithm torsional vibration control effect. BRIEF DESCRIPTION OF THE DRAWINGS

[0057] The present invention will be further described below in conjunction with the accompanying drawings:

[0058] Figure 1 The logic diagram of the torsional vibration adaptive control method for the motor drive of a planetary hybrid vehicle according to the present invention; the components represented by the reference numerals in the figure are: 1. Reconstruction of torque and feedback control; 2. Adaptive adjustment of control parameters under multiple working conditions; 3. Adaptive interference compensation;

[0059] Figure 2 It is a flow chart of the feedback control algorithm of the control method of the present invention;

[0060] Figure 3 Identification of the torque demand location of the control method of the present invention;

[0061] Figure 4 This is the INMO algorithm control principle of the control method described in the present invention. DETAILED DESCRIPTION

[0062] The present invention is described in detail below in conjunction with the accompanying drawings:

[0063] The present invention discloses a method for adaptive control of torsional vibration of motor drive of a planetary hybrid vehicle. In order to suppress the torsional vibration of the vehicle, a control algorithm is used to optimize the motor torque. The method receives the initial motor torque demand obtained by the energy management algorithm, processes it, and outputs the total target torque of the motor to the controlled vehicle. The basic principle of the method is the theory of non-cophase superposition. At the same time, the control parameters are automatically adjusted according to the changes in the working conditions, and then the interference adaptive compensation is performed. Its core includes reconstructed torque and feedback control, adaptive adjustment of multi-working condition control parameters, and adaptive interference compensation. The control method is applied to planetary multi-speed hybrid vehicles. On the basis of the energy management strategy, a set of motor torque adaptive control functions is added. The economy, power, and comfort of the vehicle can be improved without excessive modification of the original system. Next, the idea of ​​realizing adaptive torsional vibration control of motor drive of planetary hybrid vehicles is specifically explained.

[0064] See also Figure 1 The invention discloses a method for adaptively controlling torsional vibration of a motor-driven planetary hybrid vehicle, including reconstructed torque and feedback control, adaptive adjustment of control parameters under multiple working conditions, and adaptive compensation of disturbances. The method for adaptively controlling torsional vibration of a motor-driven planetary hybrid vehicle is described in detail below.

[0065] The following steps are involved:

[0066] The first step is torque reconstruction and feedback control ( Figure 1 (the position shown in ①)

[0067] like Figure 1 First, the rate of change of the initial motor torque demand is corrected through the first-order inertia link to ensure that the oscillation amplitude of the longitudinal acceleration is reduced without affecting the vehicle's dynamics, and the motor reference torque is smoothed; then the motor reference torque is divided into three parts according to the ratio of the split coefficient, and two of the parts are delayed. The three partial torques can be superimposed by using the principle of superposition of non-cophase signals to reduce system vibration to obtain the reconstructed torque of the motor; the longitudinal acceleration calculated by the nominal model is used as the feedback quantity, and the difference between the real-time longitudinal acceleration and the expected longitudinal acceleration is calculated. When the real-time longitudinal acceleration reaches the expected total acceleration, the feedback torque is output, such as Figure 2 .

[0068] The second step is to build the working condition control parameter library offline ( Figure 1 (indicated by ② in the middle)

[0069] like Figure 3 In order to improve the control effect of the torque reconstruction and feedback control algorithm, its control parameters must also be updated adaptively to track the changes in the motor torque and achieve the purpose of reducing torsional vibration. Combined with the actual vehicle road test data, the torque change time and change difference are divided into 4*4 16 working conditions at equal intervals to cover the torque change range of the vehicle during actual driving; the particle swarm intelligent algorithm is used to optimize the torque reconstruction and feedback control parameters offline. The particle value is the control parameter, which mainly includes the split coefficient, delay factor, and feedback gain; the current particle control parameters are input into the transmission system model to calculate the response, and the normalized maximum amplitude of the vehicle acceleration and the maximum amplitude of the motor-wheel equivalent speed difference are used as the fitness function of the particle swarm algorithm J obj The optimization goal

[0070]

[0071] In the formula, X is the control parameter

[0072] a diff (X)——Maximum amplitude of longitudinal acceleration

[0073] ω diff (X)——Maximum amplitude of motor-wheel equivalent speed difference

[0074] a diff_ori ——The maximum amplitude of longitudinal acceleration corresponding to no control

[0075] ω diff_ori——The maximum amplitude of longitudinal acceleration corresponding to no control The maximum amplitude of the motor-wheel equivalent speed difference

[0076] The individual optimal position and the global optimal position are recorded according to the size of the fitness; if the number of iterations is lower than the set maximum number of iterations and the fitness value is greater than the minimum fitness threshold, the speed and position of each particle are updated; when the number of iterations or the fitness value is less than the minimum fitness threshold, that is, when the termination condition is reached, the particle swarm optimization process ends and the optimal control parameter combination is output; the mapping relationship between the torque demand and the optimal control parameters is used to construct a working condition control parameter library.

[0077] The third step is to update the control parameters online adaptively ( Figure 1 (indicated by ② in the middle)

[0078] The randomly input motor torque demand is used to identify the torque location point and calculate the weighted parameters, which are then loaded into the control parameter library for online calling of the optimal control parameters.

[0079] Step 4: Interference adaptive compensation ( Figure 1 (the position shown in ③)

[0080] like Figure 4 , the INMO algorithm is used for adaptive interference compensation. The transmission shaft speed fed back by the controlled vehicle is used as input. The interference is observed based on the physical realization of the nominal inverse model and the filter design, and the interference compensation amount is calculated to reduce the influence of uncertainty interference on the torsional vibration control effect; the six-degree-of-freedom torsional vibration dynamics model is simplified to derive the nominal model transfer function matrix, and the transmission system is equivalent to the ring gear-reducer-differential and the two inertia elements of the whole vehicle. The two inertia elements are connected by an elastic element with equivalent stiffness and equivalent damping. The dynamic model of the transmission system is

[0081]

[0082] Where, T AMT_out ——Gearbox output torque

[0083] T 11 ——Transmitted torque between two inertia elements

[0084] i0——speed ratio of main reducer

[0085] I 11 ——Equivalent moment of inertia of ring gear-reducer-differential inertia components

[0086] I 12 ——Equivalent moment of inertia of the vehicle’s inertial components

[0087] ——The angle of equivalent moment of inertia of ring gear-reducer-differential

[0088] ——The angle of equivalent moment of inertia of the vehicle

[0089] c 11 ——Equivalent damping of the elastic element between two inertial elements

[0090] k 11 ——Equivalent stiffness of the elastic element between the two inertial elements

[0091] T L1 ——The sum of driving resistance torque and braking torque

[0092] Write the model in standard form of state-space equations

[0093]

[0094] By performing Lapalus transformation on the state space equation, the transfer function matrix of the system can be calculated as

[0095] P(s)=C[sI-A] -1 B+D (4)

[0096] The transfer function of the nominal model can be expressed as

[0097]

[0098] According to the Mason gain formula, the transfer function from each input to output can be calculated as

[0099]

[0100] Where P a (s)——Adaptive torque transfer function

[0101] P b (s)——disturbance transfer function

[0102] P c (s)——Noise transfer function

[0103] P act (s)——transfer function of the true model

[0104] After obtaining the transfer function of the nominal model, it is necessary to introduce a filter after the interference estimation. The filter design principle can be summarized as follows: Q(s)≈1 should be made in the low-frequency range, and Q(s)≈0 should be made in the high-frequency range, such as the Butterworth filter.

[0105] In the third step of the present invention, location point identification refers to and weighted calculation refers to:

[0106] like Figure 3, by comparing the set 16 working conditions with the randomly input motor torque change time and torque change difference, 9 areas are formed; any input torque demand will fall into the 9 areas, and then the torque change time ratio and torque change difference ratio can be directly calculated to obtain the control output u adapt .

[0107]

[0108] In the formula, K req ——Weighting factor

[0109] Δt1, Δt2, ΔT req1 , ΔT req2 ——Set the torque change time and difference required by the working condition

[0110] Δt, ΔT req ——Input at any time

[0111] u 11 、u 21 、u 12 、u 22 ——The torque time variation values ​​corresponding to the lower left, lower right, upper left and upper right of the current area.

Claims

1. A planetary hybrid vehicle motor drive torsional vibration adaptive control method, characterized in that The following steps are involved: Step 1: Torque reconstruction and feedback control First, the rate of change of the initial motor torque demand is corrected through the first-order inertia link to ensure that the oscillation amplitude of the longitudinal acceleration is reduced without affecting the vehicle's dynamics, and the motor reference torque is smoothed; then the motor reference torque is divided into three parts according to the ratio of the split coefficient, and two of the parts are delayed. The three partial torques can be superimposed by using the principle of superposition of non-cophase signals to reduce system vibration to obtain the reconstructed torque of the motor; the longitudinal acceleration calculated by the nominal model is used as the feedback quantity, and the difference between the real-time longitudinal acceleration and the expected longitudinal acceleration is calculated. When the real-time longitudinal acceleration reaches the expected total acceleration, the feedback torque is output; The second step is to build the working condition control parameter library offline Combined with the actual vehicle road test data, the torque change time and change difference are divided into 4*4 16 working conditions at equal intervals, covering the torque change range of the vehicle during actual driving; the particle swarm intelligent algorithm is used to reconstruct the torque and optimize the feedback control parameters offline. The particle value is the control parameter, which mainly includes the split coefficient, delay factor, and feedback gain; the current particle control parameters are input into the transmission system model to calculate the response, and the normalized maximum amplitude of vehicle acceleration and the maximum amplitude of the motor-wheel equivalent speed difference are used as the fitness function of the particle swarm algorithm. obj The optimization goal In the formula, X is the control parameter a diff (X)——Maximum amplitude of longitudinal acceleration ω diff (X)——Maximum amplitude of motor-wheel equivalent speed difference a diff_ori ——The maximum amplitude of longitudinal acceleration corresponding to no control ω diff_ori ——The maximum amplitude of longitudinal acceleration corresponding to no control The maximum amplitude of the motor-wheel equivalent speed difference The individual optimal position and the global optimal position are recorded according to the size of the fitness; if the number of iterations is lower than the set maximum number of iterations and the fitness value is greater than the minimum fitness threshold, the speed and position of each particle are updated; when the number of iterations or the fitness value is less than the minimum fitness threshold, that is, when the termination condition is reached, the particle swarm optimization process ends and the optimal control parameter combination is output; the mapping relationship between the torque demand and the optimal control parameters is used to construct a working condition control parameter library; The third step is to adaptively update the control parameters online. The randomly input motor torque demand is used to identify the torque location and calculate the weighted parameters, which are then loaded into the control parameter library to call the optimal control parameters online. Step 4: Interference adaptive compensation The Inverse Nominal Model Observer (INMO) algorithm based on the nominal inverse model is used for adaptive disturbance compensation. The transmission shaft speed fed back by the controlled vehicle is used as input. The disturbance is observed based on the physical implementation and filter design of the nominal inverse model, and the disturbance compensation amount is calculated to reduce the influence of uncertainty disturbance on the torsional vibration control effect. The six-degree-of-freedom torsional vibration dynamics model is simplified to derive the nominal model transfer function matrix, and the transmission system is equivalent to two inertia elements of the ring gear-reducer-differential and the whole vehicle. The two inertia elements are connected by an elastic element with equivalent stiffness and equivalent damping. The dynamic model of the transmission system is: Where, T AMT_out ——Gearbox output torque T 11 ——Transmitted torque between two inertia elements i0——speed ratio of main reducer I 11 ——Equivalent moment of inertia of ring gear-reducer-differential inertia components I 12 ——Equivalent moment of inertia of the vehicle’s inertial components ——The angle of equivalent moment of inertia of ring gear-reducer-differential ——The angle of equivalent moment of inertia of the vehicle c 11 ——Equivalent damping of the elastic element between two inertial elements k 11 ——Equivalent stiffness of the elastic element between the two inertial elements T L1 ——The sum of driving resistance torque and braking torque Write the model in standard form of state-space equations By performing Lapalus transformation on the state space equation, the transfer function matrix of the system can be calculated as P(s)=C[sI-A] -1 B+D (4) The transfer function of the nominal model can be expressed as According to the Mason gain formula, the transfer function from each input to output can be calculated as Where P a (s)——Adaptive torque transfer function P b (s)——disturbance transfer function P c (s)——Noise transfer function P act (s)——transfer function of the true model After obtaining the transfer function of the nominal model, it is necessary to introduce a filter after the interference estimation. The filter design principle can be summarized as follows: Q(s)≈1 should be achieved in the low-frequency range, and Q(s)≈0 should be achieved in the high-frequency range.

2. A planetary hybrid vehicle motor drive torsional vibration adaptive control method according to claim 1, characterized in that: In the third step, location point identification and weighted calculation refer to: By comparing the set 16 working conditions with the randomly input motor torque change time and torque change difference, 9 areas are formed; any input torque demand will fall into the 9 areas, and then the torque change time ratio and torque change difference ratio can be directly calculated to obtain the control output u adapt ; In the formula, K req ——Weighting factor Δt1, Δt2, ΔT req1 , ΔT req2 ——Set the torque change time and difference required by the working condition Δt, ΔT req ——Input at any time u 11 、u 21 、u 12 、u 22 ——The torque time variation values ​​corresponding to the lower left, lower right, upper left and upper right of the current area.

Images