Electromagnetic thermal dual-coupling temperature rise prediction and dual-cycle optimization method for active suspension actuator

Abstract

The invention discloses an electromagnetic thermal dual-coupling temperature rise prediction and dual-loop optimization method for an active suspension actuator, and the method comprises the steps: building an electric-magnetic-thermal bidirectional coupling relation transfer model of an active suspension system actuator; iteratively solving a steady-state / transient-state temperature distribution function of each lattice point of the actuator based on a mapping relation between pavement excitation and linear thrust of the actuator, and obtaining an electro-magnetic-thermal coupling characteristic of the actuator under each iteration generation number; wherein the steady-state / transient-state temperature distribution function is used for temperature rise prediction of the active suspension actuator; the dual-loop optimization strategy of the active suspension system is constructed by taking the electro-magnetic-thermal coupling performance of the actuator as an optimization target and taking the vehicle suspension performance as a constraint condition, and the optimal parameters required during the design of the active suspension system are obtained. According to the design parameters obtained through the method, the design efficiency and effect of the active suspension system can be improved, and the designed active suspension system has the advantages of being low in energy consumption, high in control precision and good in reliability.

Description

Technical Field

[0001] The invention relates to the technical field of active suspension, and in particular to an electromagnetic-thermal dual-coupling temperature rise prediction and dual-loop optimization method for an active suspension actuator. Background Art

[0002] At present, in the research related to automotive active suspension actuators, the electromagnetic properties of the actuators are mainly studied, and few scholars have studied the working thermal characteristics of the actuators, resulting in the inability to provide accurate guidance in the actuator design optimization link. In addition, most of the existing studies are based on the establishment of analytical equations based on the two-body two-degree-of-freedom active suspension structure, which is quite different from the actual vehicle suspension structure, resulting in a large error between the suspension dynamics state variables in the calculation results and the actual physical situation, which is not conducive to the prediction of the temperature rise of the actuator and the development of related technologies. Therefore, it is very necessary to consider the electro-magnetic-thermal bidirectional coupling effect of the actuator, predict the temperature rise, and establish a more accurate active suspension model.

[0003] The working efficiency, reliability and main force control accuracy of the automotive active suspension system are particularly important performances. As the core component, the actuator, the less electromagnetic loss it generates during operation, the higher the system efficiency; the actuator is accompanied by temperature rise during operation. If the actuator is at a high temperature working point for a long time, it is prone to risks such as structural deformation and insulation failure, reducing the reliability of the system; when the actuator is working under load, due to the influence of ripple disturbance, tooth effect, edge effect, friction disturbance, and current delay harmonics, electromagnetic thrust fluctuations will inevitably occur, resulting in a decrease in the system's control accuracy over the main force. Therefore, in view of the impact of the negative effects of motor operation on system performance, the electromagnetic thermal coupling performance of the actuator needs to be further optimized, aiming to break through the technical bottleneck of low efficiency and poor effect of the optimization design of the actuator-type active suspension system, and solve the problems of high energy consumption, low control accuracy and poor reliability of the electromagnetic actuator-type active suspension system. Summary of the invention

[0004] In view of this, the present invention provides an electromagnetic-thermal dual-coupled temperature rise prediction and dual-loop optimization method for an active suspension actuator.

[0005] The present invention achieves the above technical objectives through the following technical means.

[0006] Electromagnetic-thermal dual-coupling temperature rise prediction and dual-loop optimization method for active suspension actuators:

[0007] Establishing road surface excitation Z based on BP neural network algorithm r With actuator linear thrust F d The mapping relationship;

[0008] The electro-magnetic-thermal bidirectional coupling transfer model of the active suspension system actuator is established based on the road surface excitation Z rWith actuator linear thrust F d The steady-state / transient temperature distribution function of each point of the actuator is iteratively solved according to the mapping relationship, and the electro-magnetic-thermal coupling characteristics of the actuator under each iterative algebra are obtained; the steady-state / transient temperature distribution function is used to predict the temperature rise of the active suspension actuator;

[0009] Taking the electro-magnetic-thermal coupling performance of the actuator as the optimization target and the vehicle suspension performance as the constraint condition, a double-loop optimization strategy for the active suspension system is constructed to obtain the optimal parameters required for the design of the active suspension system.

[0010] The electromagnetic thermal coupling performance of the actuator includes thrust fluctuation F b and maximum temperature rise T g The vehicle suspension performance includes dynamic travel, vehicle body acceleration and root mean square value of wheel dynamic load.

[0011] Furthermore, the electric-magnetic-thermal bidirectional coupling relationship transfer model of the active suspension system actuator is composed of six units arranged and connected: the first unit is a force-electric coupling transfer unit I i (F d , P εi ), participates in the electro-magnetic-thermal forward coupling and reverse coupling process; the second unit is the electro-magnetic coupling transfer unit P i (I abci ), only participates in the electro-magnetic-thermal forward coupling process; the third unit is the magnetic-thermal coupling transfer unit T i (P Qi ), only participates in the electro-magnetic-thermal forward coupling process; the fourth unit is the thermal-magnetic coupling transfer unit P i+1 (T xyi ), only participates in the electric-magnetic-thermal reverse coupling process; the fifth unit is the thermal-electric coupling transfer unit I i+1 (T xyi ), only participates in the electro-magnetic-thermal reverse coupling process; the sixth unit is the thermal-thermal coupling transfer unit T i+2 (T xyi+1 ), at the critical point of the electro-magnetic-thermal forward and reverse coupling process, the temperature distribution in the temperature field is corrected; where: I i is the set of electric field physical parameters for iterative calculation of the i-th generation, P i is the set of physical parameters of the magnetic field during the iterative calculation of the i-th generation, T i is the set of thermal field physical parameters for iterative calculation of generation i, i = 0, 1, 2…N.

[0012] Furthermore, the road surface excitation Z r With actuator linear thrust F d The steady-state temperature distribution function of each point of the actuator is iteratively solved by the mapping relationship, which is:

[0013] Based on the mapping relationship, when the vehicle road grade and vehicle speed are determined, the actuator thrust F within the road excitation action time Δt is obtained. d , as the first generation I0(F d , P ε0 ) input item, through F d and the permanent magnet flux physical parameter P ε0 , establish the initial characteristic set I0 of the electric field, and transform the three-phase current I abc0 As the first generation of electro-magnetic coupling transfer unit abc0 ) input item, establish a set P0 representing the initial characteristics of the magnetic field, and convert the electromagnetic loss physical parameter P in P0 Q0 The first generation T0(P) as heat source input magneto-thermal coupling transfer unit Q0 ), establish the initial characteristic set T0 to characterize the thermal field, and transform the temperature distribution T in T0 xy0 As thermal-magnetic coupling transfer unit P1 (T xy0 )、Thermo-electric coupling transfer unit I1(T xy0 ) input items, establish the next generation characteristic set P1 representing the magnetic field and the next generation characteristic set I1 representing the electric field, repeat the above steps to calculate the electric, magnetic and thermal field characteristics;

[0014] In the calculation process of algebra i=1,2…N, the physical parameter T in the thermal field characteristic set xyi , which is different from the direct input of the thermal-magnetic coupling transfer unit P when algebraic i = 0 i+1 (T xyi ) and thermal-electric coupling transfer unit I i+1 (T xyi ), but first as the input item of the thermal-thermal coupling transfer unit, to establish the next generation characteristic set T of the thermal field i+1 , then the T i+1 The temperature distribution in T xyi As input to the thermal-magnetic coupling transfer unit and the thermal-electrical coupling transfer unit.

[0015] Furthermore, the determination of the electric field characteristic set I1 is subject to the force-electric coupling transfer unit I1 (F d , P ε1 ) parameter P ε1 The influence of I1(F d , P ε1 ) has a transitive relationship of In order to maintain the active force required by the system, that is, the actuator thrust F d , it is necessary to increase the three-phase current in I1, which is transmitted through the transfer relationship i d 、i qTransformation is obtained; where τ is the magnetic pole spacing of the actuator, L d and L q are the d-axis and q-axis inductances, respectively, and i d and i q are the d and q axis currents respectively.

[0016] Furthermore, the convergence condition of the iteration is: when the temperature distribution difference of each point in three consecutive generations is less than 0.2°C, the steady-state temperature distribution function of the actuator under the road surface excitation Δt is obtained.

[0017] Furthermore, the electro-magnetic-thermal coupling performance optimization function of the actuator is:

[0018]

[0019] And the following constraints are met:

[0020]

[0021] Where W is the actuator structure parameter set, j is the large cycle iteration number, σ1 is the body acceleration weight, σ2 is the tire dynamic load weight, rms_s is the RMS value of the suspension dynamic travel, rms_b is the RMS value of the body acceleration, rms_d is the RMS value of the tire dynamic load, and a and b are constants.

[0022] Furthermore, the dual-loop optimization strategy of the active suspension system specifically includes:

[0023] Take the large loop iteration number j = 0, the small loop iteration number i = 0, and the change amount of each structural parameter of the actuator Changes in electro-magnetic-thermal coupling performance The quantitative influence degree is recorded as η. Based on η, the discrete interval z / η of the structural parameters of the Taguchi method experiment is obtained, and the sensitivity of each structural parameter of the actuator to the thrust fluctuation and the maximum temperature rise is calculated. Select sensitivity The structural parameters in the range of greater than 0.5 and less than -0.5 were fitted with response surface functions to establish the functional relationship between the highly sensitive structural parameters and the electromagnetic-thermal coupling performance; based on the functional relationship, the genetic algorithm was used to obtain the electro-magnetic-thermal coupling performance (T g ) j 、(F b ) j The Pareto frontier curve of the jth generation is obtained and recorded [w1, w2, w3...] j and the jth generation magnetothermal coupling performance (T g ) j 、(F b ) j, and judge whether the elements in the control coefficient combination set U are all taken. If the control coefficient combination set U is not taken, continue to cycle the above process, that is, j+1, until all the control coefficient combinations are taken, and compare the optimal values ​​of each generation. The smallest corresponding control parameter combination and actuator structure parameters are the global optimal parameter set, which are used as the design parameters of the active suspension system;

[0024] in, z is the structural size ratio coefficient.

[0025] Furthermore, the process of obtaining the control coefficient combination set U is:

[0026] Based on the numerical analytical equation of the five-body two-degree-of-freedom active suspension, the ceiling damping constant coefficient C is obtained. sky , Passive damping constant coefficient C s The mapping relationship between the combination of and the root mean square value of suspension dynamic travel rms_s, the root mean square value of body acceleration rms_b, and the root mean square value of tire dynamic load rms_d is obtained, and then the constraint condition cloud diagram is obtained. In the constraint condition area that satisfies the electro-magnetic-thermal coupling performance optimization function, C sky , C s Take a set element at intervals to obtain a control parameter combination set U that meets the constraint conditions.

[0027] Furthermore, in order to obtain the change of each structural parameter of the actuator Changes in electro-magnetic-thermal coupling performance The quantitative influence of the actuator initial structural parameter combination W0 is obtained under the steady-state electro-magnetic-thermal field coupling characteristic set {I j , P j 、T j}, continue to obtain the steady-state electric-magnetic-thermal field coupling characteristic set under other structural parameter combinations in the actuator structural parameter combination set W, and calculate different actuator structural parameter combinations. It is necessary to modify the force-electric coupling transfer unit in the electric-magnetic-thermal bidirectional coupling relationship transfer model of the active suspension system actuator. The numerical calculation equation in the transfer unit is:

[0028]

[0029] Where τ is the magnetic pole spacing of the actuator, L d and L q are the d-axis and q-axis inductances, respectively, and i d and i q are d and q axis currents respectively;

[0030] Changes in actuator structural parameters will change τ, P, L d , L q , due to F d It is known that, therefore, id and i q The value of changes, realizing the correction of the force-electric coupling transmission relationship.

[0031] Furthermore, the numerical analytical equation of the five-body two-degree-of-freedom active suspension includes a comprehensive dynamic equation of the nonlinear five-body two-degree-of-freedom active suspension:

[0032]

[0033] Among them, Q j is the generalized external force, T is the overall kinetic energy of the five-body two-degree-of-freedom active suspension, q j is the generalized coordinate, V is the overall potential energy of the five-body two-degree-of-freedom active suspension, is the force on the ith particle, is the potential vector of the i-th particle, F ix is the component of the force on the ith particle on the x-axis, F iy is the component of the force on the ith particle on the y-axis, x i and i is the x and y axis coordinates of point i, and n is a constant;

[0034] The overall kinetic energy of the five-body two-degree-of-freedom active suspension includes the kinetic energy of the tires. Among them, m u For tire quality, is the velocity of the hinge point A, is the angle change between the composite structure axis and the vertical direction;

[0035] Moment of inertia of composite structure AB in five-body two-degree-of-freedom active suspension in, are the mass of composite component 2 in composite structure AB, the mass of composite component 1 in composite structure AB, the length of composite component 1, and the length of composite component 2, respectively. AN is the distance between the hinge point A and the center of mass N of the composite structure.

[0036] The beneficial effects of the present invention are:

[0037] (1) The present invention establishes a numerical analytical equation for a five-body two-degree-of-freedom active suspension. Compared with the traditional analytical equation for active suspension, the present invention retains the jumping and lateral movement characteristics of real wheels, takes into account the influence of the topological rod connection relationship and the mass of the topological rod on kinematics and dynamics, and greatly improves the calculation accuracy of the performance parameters of the active suspension system and the operating parameters of the actuator.

[0038] (2) The electric-magnetic-thermal bidirectional coupling relationship transfer model of the active suspension actuator of the present invention takes into account the influence of temperature on the permanent magnet flux, winding resistivity, material magnetic permeability and material thermal conductivity. It can calculate the number of iterations and the electric field coupling characteristics, magnetic field coupling characteristics and thermal field coupling characteristics of the active suspension actuator under the convergence conditions. In addition, it can calculate the transient thermal distribution and steady-state thermal distribution of the active suspension actuator.

[0039] (3) The present invention takes the electro-magnetic-thermal coupling performance of the actuator as the optimization target and the vehicle suspension performance as the constraint condition to construct a dual-loop optimization strategy for the active suspension system, thereby obtaining the optimal parameters required for the design of the active suspension system and improving the efficiency and effect of the optimization design of the actuator-type active suspension system. The active suspension system designed with the parameters obtained by the dual-loop optimization strategy has low energy consumption, high control accuracy and good reliability. BRIEF DESCRIPTION OF THE DRAWINGS

[0040] Figure 1 A block diagram of the electromagnetic-thermal dual-coupling temperature rise prediction and dual-loop optimization method for the active suspension actuator of the present invention;

[0041] Figure 2 It is a simplified structural diagram of the five-body two-degree-of-freedom active suspension of the present invention;

[0042] Figure 3 The road surface excitation Z is established according to the present invention. r and actuator thrust F d Schematic diagram of the mapping relationship;

[0043] Figure 4 It is an iterative process diagram of the electrical, magnetic and thermal field physical parameters of the present invention;

[0044] FIG5( a ) is a schematic diagram of contour lines of a control parameter set that meets constraint condition 1 according to the present invention;

[0045] FIG5( b ) is a schematic diagram of contour lines of a control parameter set that meets constraint condition 2 according to the present invention;

[0046] Figure 6 This is a schematic diagram of a region satisfying two constraints according to the present invention;

[0047] Figure 7 This is a flow chart of the dual-loop optimization strategy described in the present invention. DETAILED DESCRIPTION

[0048] The present invention is further described below in conjunction with the accompanying drawings and specific embodiments, but the protection scope of the present invention is not limited thereto.

[0049] See also Figure 1This embodiment provides an electromagnetic-thermal dual-coupling temperature rise prediction and dual-loop optimization method for an active suspension actuator, which specifically includes the following contents:

[0050] S1, establish the numerical analytical equation of the five-body two-degree-of-freedom active suspension, and verify the accuracy of the analytical equation through the finite element model.

[0051] Based on the topology of double wishbone suspension, the following Figure 2 A nonlinear five-body two-degree-of-freedom active suspension structure is shown, and then the numerical analytical equation of the five-body two-degree-of-freedom active suspension is established in MATLAB, including:

[0052] Random road excitation function Where u is the instantaneous speed of the vehicle with a double wishbone suspension topology, G q (n0) is the grade constant of the road surface on which the vehicle is traveling, n0 is the reference spatial frequency, and ω(t) represents Gaussian white noise;

[0053] Mixed Shed Control Strategy Among them, F d is the instantaneous main force of the active suspension actuator, the ceiling damping constant coefficient C sky and the passive damping constant coefficient C s is the active suspension control parameter, z u is the vertical speed of the tire, z s is the vertical velocity of the vehicle body;

[0054] The comprehensive dynamic equation of nonlinear five-body two-degree-of-freedom active suspension is:

[0055]

[0056] Among them, Q j is the generalized external force, T is the overall kinetic energy of the five-body two-degree-of-freedom active suspension, q j is the generalized coordinate, V is the overall potential energy of the five-body two-degree-of-freedom active suspension, is the force on the ith particle, is the potential vector of the i-th particle, F ix is the component of the force on the ith particle on the x-axis, F iy is the component of the force on the ith particle on the y-axis, x i and i are the x and y axis coordinates of point i, and n is a constant.

[0057] The above comprehensive dynamic equation is composed of the following sub-equations:

[0058] The kinetic energy expression of the vehicle body OA is:

[0059]

[0060] The kinetic energy expression of the lower swing arm OB:

[0061]

[0062] The kinetic energy expression of the composite structure AB is:

[0063]

[0064] The kinetic energy of the tire is expressed as:

[0065]

[0066] The overall kinetic energy expression of the five-body two-degree-of-freedom active suspension is:

[0067] T=T OA +T OB +T AB +T wheel

[0068] The elastic potential energy expression of the suspension spring is:

[0069]

[0070] The elastic potential energy expression of the tire is:

[0071]

[0072] The gravitational potential energy equation of the five-body two-degree-of-freedom active suspension is expressed as follows:

[0073]

[0074] The generalized external force expression of the five-body two-degree-of-freedom active suspension is:

[0075]

[0076] Among them, m OA 、m OB 、m AB 、m u They are the body mass, lower arm mass, composite structure mass, and tire mass. OB , L AB , L AB0 , L AN , L OM , H s They are the length of the lower arm, the original length of the composite component, the change in the length of the composite component, the distance between the hinge point A and the center of mass of the composite structure N, the distance between the hinge point O and the center of mass of the lower arm M, the distance between the hinge point A and the center of mass of the vehicle body C ms The vertical distance, k s , kt are the coil spring stiffness and tire stiffness respectively. are the initial angle between the composite structure axis and the vertical direction, and the angle change value, respectively. θ0 and θ are the initial angle and angle change between the axis of the lower arm and the horizontal direction, respectively, which belong to the suspension parameters; is the velocity at point O, with vertical upwards being the positive direction; is the angular velocity of the lower swing arm, counterclockwise is the positive direction; is the velocity at point A, with vertical upwards being the positive direction; Z s , Z u , Z r They are body displacement, tire bounce displacement, and road excitation; KG AB , KG OA , KG OB , KG wheel They are the gravitational potential energy of the composite structure AB, the gravitational potential energy of the vehicle body OA, the gravitational potential energy of the lower arm OB, and the gravitational potential energy of the tire; is the main force from point A1 to the positive direction of A, is the main force from point B1 to the positive direction of B, is the position vector of point A1, is the position vector of point B1.

[0077] L AB , J N They satisfy the following relationships respectively:

[0078]

[0079] Among them, J N is the moment of inertia of the composite structure AB, m A1B1 、m AA1 , L AA1 , L A1B1 are the mass of composite component 2, the mass of composite component 1, the length of composite component 1, and the length of composite component 2 respectively; α and β are the angles between the line connecting the hinge point OA and the horizontal direction and the angle between the line connecting the hinge point OA and the vertical direction respectively, L OA is the distance between the hinge points O and A, L AN is the distance between the hinge point A and the center of mass N of the composite structure.

[0080] Based on modern virtual prototype technology, secondary development is carried out on the basis of ADAMS program. ADAMS View is used to build an active suspension finite element dynamic model that is consistent with the performance parameters, structural parameters and control strategy of the above-mentioned five-body two-degree-of-freedom active suspension. Under the same road excitation conditions, the dynamic feature set of the analytical model (i.e., numerical analytical equation) and the dynamic feature set of the finite element method model are captured, and the dynamic feature sets (body acceleration, suspension dynamic travel, wheel dynamic load, active force) are compared to verify the accuracy of the numerical analytical equation.

[0081] S2, establish road surface incentive Z based on BP neural network algorithm r With actuator linear thrust F d The mapping relationship.

[0082] According to the relevant theoretical knowledge of mechanical freedom, it can be known that the five-body two-degree-of-freedom active suspension model requires two external inputs to determine the unique motion state of the machine. Since the five-body two-degree-of-freedom active suspension system establishes a coupling relationship between the two degrees of freedom through the coil spring, the road excitation Z r The motion characteristics and dynamic characteristics of the active suspension system can be determined. The present invention establishes the road excitation Z based on the BP neural network algorithm. r and actuator thrust F d The mapping relationship of BP network establishment process is shown in Figure 3 , specifically:

[0083] S21: Initialization parameters: Initialize the neural network related parameters, including neural network step size, weight, balance coefficient, exponential decay factor, and calculation cycle;

[0084] S22: Set discrete sets: Set the instantaneous speed of the vehicle u as discrete set I, with a range of [10, 20], and an interval of 0.5 m / s; the road surface grade constant Gq(n0) as discrete set II, with a range of [4 2 ×10 -6 , 32 2 ×10 -6 ], every 4 2 ×10 -6 m 3 interval;

[0085] S23: Determine the number of neural network layers and neurons: Since the input layer is relatively simple, the number of hidden layers is set to 1, including 20 neurons, so as to obtain a shorter training time;

[0086] S24: Select excitation function: select hyperbolic tangent function and purelin function as the excitation functions of the hidden layer and output layer respectively;

[0087] S25: Select training function: Batch gradient descent is used as the training function to meet the need for fast calculation of the error value of the training function, so as to better adjust the neural network weights and thresholds;

[0088] S25: Calculate the error value: the difference E between the expected output and the actual output Fd The sum of squares is taken as the error function value.

[0089] Through the process of S21-S25, the road surface excitation Z is established. r and actuator thrust F d The mapping relationship.

[0090] S3, establish the electro-magnetic-thermal bidirectional coupling relationship transfer model of the active suspension system actuator, iteratively solve the steady-state / transient temperature distribution function of each node of the actuator, and obtain the electro-magnetic-thermal coupling characteristics of the actuator under each iterative algebra.

[0091] Considering the influence of temperature on permanent magnet flux, winding resistivity, material magnetic permeability and material thermal conductivity, six units are arranged and connected to form the electric-magnetic-thermal bidirectional coupling relationship transfer model of the active suspension actuator; the first unit is the force-electric coupling transfer unit I i (F d , P εi ), this unit participates in the electro-magnetic-thermal forward coupling and reverse coupling process; the second unit is the electro-magnetic coupling transfer unit P i (I abci ), this unit only participates in the electro-magnetic-thermal forward coupling process; the third unit is the magnetic-thermal coupling transfer unit T i (P Qi ), this unit only participates in the electro-magnetic-thermal forward coupling process; the fourth unit is the thermal-magnetic coupling transfer unit P i+1 (T xyi ), this unit only participates in the electric-magnetic-thermal reverse coupling process; the fifth unit is the thermal-electric coupling transfer unit I i+1 (T xyi ), this unit only participates in the electro-magnetic-thermal reverse coupling process; the sixth unit is the thermal-thermal coupling transfer unit T i+2 (T xyi+1 ), this unit is at the critical point of the electro-magnetic-thermal forward and reverse coupling process, and the temperature distribution in the temperature field is corrected. Among them: I i is the set of electric field physical parameters (electric field characteristics) when iteratively calculating the i-th generation (i = 0, 1, 2 ... N), P i is the set of physical parameters (magnetic field characteristics) of the magnetic field when iteratively calculating the i-th generation (i=0,1,2…N), T i It is the set of thermal field physical parameters (thermal field characteristics) during iterative calculation of generation i (i=0,1,2…N).

[0092] See also Figure 4 The iterative process of the electrical, magnetic and thermal field physical parameters is as follows: Based on the mapping relationship established in step S2, when the vehicle road surface grade and vehicle speed are determined, the actuator thrust F within the road surface excitation time Δt is obtained. d , as the first unit of force-electric coupling transmission unit of the first generation I0 (F d , P ε0 ) input item, through F d The physical parameter P of the permanent magnet flux in the magnetic field characteristics ε0 (initial magnetic field), establish the initial characteristic set I0 to characterize the electric field, and transform the three-phase current I abc0 As the first generation P0(I) of the second unit - electro-magnetic coupling transfer unit abc0 ) input item, establish the initial characteristic set P0 of the magnetic field, and convert the electromagnetic loss physical parameter P in the initial characteristic set P0 of the magnetic field into Q0 As the heat source input to the third unit - the first generation T0 (P Q0 ), establish the initial characteristic set T0 to characterize the thermal field, and convert the temperature distribution T in the initial thermal field characteristic set T0 xy0 As the fourth unit - thermal-magnetic coupling transfer unit P1 (T xy0 ), the fifth unit - thermal-electric coupling transfer unit I1 (T xy0 ) input items, establish the next generation characteristic set P1 representing the magnetic field and the next generation characteristic set I1 representing the electric field. The determination of the electric field characteristic set I1 is affected by the force-electric coupling transfer unit I1 (F d , P ε1 ) parameter P ε1 The influence of (I1(F d , P ε1 ) The transitive relationship is As the temperature of the permanent magnets increases, the magnetic flux of each permanent magnet decreases. In order to maintain the main force required by the system, that is, the actuator thrust F d , it is necessary to increase the three-phase current in I1, which is transferred through the relationship i d 、i q Transformed), repeat the above steps to calculate the electrical, magnetic and thermal field characteristics (i = 1, 2, 3…N); it is worth noting that the physical parameter T in the subsequent thermal field characteristic set xyi , which is different from the direct input of the thermal-magnetic coupling transfer unit P when algebraic i = 0 i+1 (T xyi ) and thermal-electric coupling transfer unit I i+1 (T xyi ), but first as the input item of the sixth unit - heat-heat coupling transfer unit, to establish the next generation characteristic set of thermal field T i+1, and then the next generation thermal field characteristic set T i+1 Temperature distribution T xyi (i≠0) as the input of the fourth unit - thermal-magnetic coupling transfer unit and the fifth unit - thermal-electric coupling transfer unit. It is worth noting that the convergence condition of the iteration is: when three consecutive generations (T xyi+2 、T xyi+1 、T xyi ) the temperature distribution difference of each point is less than 0.2℃, that is, the steady-state temperature distribution function T of the actuator under the road surface excitation (duration Δt) is obtained. xy When Δt is extremely small, the transient temperature distribution function of the actuator in the active suspension system can be calculated.

[0093] In the iterative process, {I0, P0, T0} is the set of electro-magnetic-thermal coupling characteristics of the first generation actuator, and so on, the electro-magnetic-thermal coupling characteristics of the actuator under each iteration generation are obtained.

[0094] S4, based on the Taguchi method-response surface method-genetic algorithm, takes the thrust fluctuation and maximum temperature rise in the electro-magnetic-thermal coupling characteristics of the actuator as optimization targets, and takes the vehicle suspension performance (dynamic stroke, body acceleration and root mean square value of wheel dynamic load) as constraints to construct a double-loop optimization strategy for the active suspension system; among which, the maximum temperature rise is determined by the steady-state temperature distribution function.

[0095] Active suspension system dual-loop optimization strategy Figure 7 As shown, according to the electromagnetic thermal coupling relationship of the actuator and the numerical analytical equation analysis of the active suspension, the maximum temperature rise T g It is not only related to the structure of the actuator itself, but also to the magnitude of the active force of the active suspension system, that is, the active suspension control parameter C sky , C s Similarly, the thrust fluctuation F b It is not only related to the actuator structural parameters, but also to the actuator maximum temperature rise. The dynamic travel of the active suspension, the body acceleration and the tire dynamic load are all related to C sky , C s The value is related. In order to ensure the performance of active suspension, C sky , C s The value of needs to be constrained.

[0096] Therefore, the optimization function of the electromagnetic thermal coupling performance (thrust fluctuation and maximum temperature rise) of the active suspension actuator can be written as:

[0097]

[0098] And the following constraints are met:

[0099]

[0100] Among them, W is the actuator structure parameter combination set, j is the large cycle iteration number, σ1 is the body acceleration weight, σ2 is the tire dynamic load weight, and the values ​​of σ1 and σ2 are determined according to the actual driving conditions of the vehicle. Generally, low-speed bumpy roads pay more attention to driving smoothness, that is, σ1 can take a larger value, and high-speed flat roads pay more attention to safety, that is, σ2 can take a larger value; rms_s is the root mean square value of the suspension dynamic travel, rms_b is the root mean square value of the body acceleration, rms_d is the root mean square value of the tire dynamic load, and the values ​​of constants a and b are related to the vehicle model and are empirical values.

[0101] Based on the numerical analytical equation of the five-body two-degree-of-freedom active suspension, C sky (within the range of 0Ns / m-2000Ns / m), C s The mapping relationship between the combination of (within the range of 0.5Ns / m-2500Ns / m) and the root mean square value of suspension travel rms_s, the root mean square value of body acceleration rms_b, and the root mean square value of tire dynamic load rms_d is obtained, and then the constraint condition cloud diagram is obtained, which meets the constraint condition rms_s[(c sky ) j 、(c s ) j The control parameter set with ]≤a is located in the area above the contour curve shown in the figure, as shown in Figure 5(a), which meets the constraint condition 1500·σ1·rms_b[(c sky ) j 、(c s ) j ]+σ2·rms_d[(c sky ) j 、(c s ) j ]≤b is located in the area below the contour curve shown in the figure, as shown in Figure 5(b); the area that satisfies both constraints is shown in Figure 6 In this region, C sky , C s A set element is taken at every interval of 150Ns / m to obtain a control parameter combination set U that meets the two constraints.

[0102] Take the large loop iteration number j = 0 and the small loop iteration number i = 0, and according to the mixed shed control strategy of the active suspension system, obtain (F d ) j Based on the electro-magnetic-thermal bidirectional coupling relationship transfer model of the S3 active suspension system actuator, the steady-state electro-magnetic-thermal field coupling characteristic set {I j , P j 、T j}, in order to obtain the change of each structural parameter of the actuator Changes in electro-magnetic-thermal coupling performance In order to determine the quantitative influence of the actuator structure parameter combination set W, it is necessary to continue to obtain the steady-state electro-magnetic-thermal field coupling characteristic set under other structural parameter combinations in the actuator structure parameter combination set W. To calculate different actuator structure parameter combinations, it is necessary to modify the force-electric coupling transfer unit in the electro-magnetic-thermal bidirectional coupling relationship transfer model of the active suspension system actuator. The numerical calculation equation in this transfer unit is:

[0103]

[0104] Where τ is the magnetic pole spacing of the actuator, L d and L q are the d-axis and q-axis inductances, respectively, and i d and i q are d and q axis currents respectively;

[0105] Changes in actuator structural parameters will change τ, P, L d , L q , due to F d It is known that, therefore, i d and i q The value of changes, realizing the correction of the force-electric coupling transmission relationship;

[0106] The above quantitative influence degree is recorded as η, and satisfies the expression:

[0107]

[0108] Based on η, the discrete interval z / η of the structural parameters of the Taguchi method experiment is obtained, and it is written into the Minitab data processing software to calculate the sensitivity of each structural parameter of the actuator to thrust fluctuation and maximum temperature rise. (Value range: ), where z is the structural size ratio coefficient.

[0109] Select sensitivity Structural parameters in the range of greater than 0.5 and less than -0.5, eliminating insensitive parameters Perform response surface function fitting to establish the functional relationship between highly sensitive structural parameters and electromagnetic thermal coupling performance: F b (w1,w2,w3····),T g (w1,w2,w3····), where w1, w2, w3... are highly sensitive structural parameters.

[0110] Based on the established functional relationship, the genetic algorithm is used to obtain the electro-magnetic-thermal coupling performance (T g ) j 、(F b ) jThe Pareto frontier curve of the jth generation is obtained and recorded [w1, w2, w3...] j and the jth generation magnetothermal coupling performance (T g ) j 、(F b ) j The minimum value (contemporary optimal value) is used to determine whether all elements in the control coefficient combination set U have been taken. If the control coefficient combination set U has not been taken, the above process is continued to cycle, i.e. j+1, until all control coefficient combinations are taken. The optimal values ​​of each generation are compared, and the smallest corresponding control parameter combination and actuator structure parameters are the global optimal parameter set, which are used as the design parameters of the active suspension system.

[0111] The embodiments are preferred implementations of the present invention, but the present invention is not limited to the above-mentioned implementations. Any obvious improvements, substitutions or modifications that can be made by those skilled in the art without departing from the essential content of the present invention belong to the protection scope of the present invention.

Claims

1. An electromagnetic-thermal dual-coupling temperature rise prediction and dual-loop optimization method for active suspension actuators, characterized by: Establishing road surface excitation Z based on BP neural network algorithm r With actuator linear thrust F d The mapping relationship; The electro-magnetic-thermal bidirectional coupling transfer model of the active suspension system actuator is established based on the road surface excitation Z r With actuator linear thrust F d The steady-state / transient temperature distribution function of each point of the actuator is iteratively solved according to the mapping relationship, and the electro-magnetic-thermal coupling characteristics of the actuator under each iterative algebra are obtained; the steady-state / transient temperature distribution function is used to predict the temperature rise of the active suspension actuator; Taking the electro-magnetic-thermal coupling performance of the actuator as the optimization target and the vehicle suspension performance as the constraint condition, a double-loop optimization strategy for the active suspension system is constructed to obtain the optimal parameters required for the design of the active suspension system. The electromagnetic thermal coupling performance of the actuator includes thrust fluctuation F b and maximum temperature rise T g The vehicle suspension performance includes dynamic travel, vehicle body acceleration and root mean square value of wheel dynamic load.

2. The electromagnetic-thermal dual-coupling temperature rise prediction and dual-cycle optimization method according to claim 1 is characterized in that: The electric-magnetic-thermal bidirectional coupling relationship transfer model of the active suspension system actuator is composed of six units arranged and connected: the first unit is the force-electric coupling transfer unit I i (F d , P εi ), participate in the electro-magnetic-thermal forward coupling and reverse coupling process; The second unit is the electro-magnetic coupling transfer unit P i (I abci ), only participates in the electro-magnetic-thermal forward coupling process; the third unit is the magnetic-thermal coupling transfer unit T i (P Qi ), only participates in the electro-magnetic-thermal forward coupling process; the fourth unit is the thermal-magnetic coupling transfer unit P i+1 (T xyi ), only participates in the electro-magnetic-thermal reverse coupling process; The fifth unit is the thermal-electric coupling transfer unit I i+1 (T xyi ), only participates in the electro-magnetic-thermal reverse coupling process; The sixth unit is the heat-heat coupling transfer unit T i+2 (T xyi+1 ), at the critical point of the electro-magnetic-thermal forward and reverse coupling process, the temperature distribution in the temperature field is corrected; where: I i is the set of electric field physical parameters for iterative calculation of the i-th generation, P i is the set of physical parameters of the magnetic field during the iterative calculation of the i-th generation, T i is the set of thermal field physical parameters for iterative calculation of generation i, i = 0, 1, 2…N.

3. The electromagnetic-thermal dual-coupling temperature rise prediction and dual-cycle optimization method according to claim 2 is characterized in that: The road surface excitation Z r With actuator linear thrust F d The steady-state temperature distribution function of each point of the actuator is iteratively solved by the mapping relationship, which is: Based on the mapping relationship, when the vehicle road grade and vehicle speed are determined, the actuator thrust F within the road excitation action time Δt is obtained. d , as the first generation I0(F d , P ε0 ) input item, through F d and the permanent magnet flux physical parameter P ε0 , establish the initial characteristic set I0 of the electric field, and transform the three-phase current I abc0 As the first generation of electro-magnetic coupling transfer unit abc0 ) input item, establish a set P0 representing the initial characteristics of the magnetic field, and convert the electromagnetic loss physical parameter P in P0 Q0 The first generation T0(P) as heat source input magneto-thermal coupling transfer unit Q0 ), establish the initial characteristic set T0 to characterize the thermal field, and transform the temperature distribution T in T0 xy0 As thermal-magnetic coupling transfer unit P1 (T xy0 )、Thermo-electric coupling transfer unit I1(T xy0 ) input items, establish the next generation characteristic set P1 representing the magnetic field and the next generation characteristic set I1 representing the electric field, repeat the above steps to calculate the electric, magnetic and thermal field characteristics; In the calculation process of algebra i=1,2…N, the physical parameter T in the thermal field characteristic set xyi , which is different from the direct input of the thermal-magnetic coupling transfer unit P when algebraic i = 0 i+1 (T xyi ) and thermal-electric coupling transfer unit I i+1 (T xyi ), but first as the input item of the thermal-thermal coupling transfer unit, to establish the next generation characteristic set T of the thermal field i+1 , then the T i+1 The temperature distribution in T xyi As input to the thermal-magnetic coupling transfer unit and the thermal-electrical coupling transfer unit.

4. The electromagnetic-thermal dual-coupling temperature rise prediction and dual-cycle optimization method according to claim 3 is characterized in that: The determination of the electric field characteristic set I1 is subject to the force-electric coupling transfer unit I1 (F d , P ε1 ) parameter P ε1 The influence of I1(F d , P ε1 ) has a transitive relationship of In order to maintain the active force required by the system, that is, the actuator thrust F d , it is necessary to increase the three-phase current in I1, which is transmitted through the transfer relationship i d 、i q Transformation is obtained; where τ is the magnetic pole spacing of the actuator, L d and L q are the d-axis and q-axis inductances, respectively, and i d and i q are d-axis and q-axis currents respectively.

5. The electromagnetic-thermal dual-coupling temperature rise prediction and dual-cycle optimization method according to claim 3 is characterized in that: The convergence condition of the iteration is: when the temperature distribution difference of each point in three consecutive generations is less than 0.2℃, the steady-state temperature distribution function of the actuator under the road surface excitation Δt is obtained.

6. The electromagnetic-thermal dual-coupling temperature rise prediction and dual-cycle optimization method according to claim 1 is characterized in that: The electro-magnetic-thermal coupling performance optimization function of the actuator is: And the following constraints are met: Where W is the actuator structure parameter set, j is the large cycle iteration number, σ1 is the body acceleration weight, σ2 is the tire dynamic load weight, rms_s is the RMS value of the suspension dynamic travel, rms_b is the RMS value of the body acceleration, rms_d is the RMS value of the tire dynamic load, and a and b are constants.

7. The electromagnetic-thermal dual-coupling temperature rise prediction and dual-cycle optimization method according to claim 1 is characterized in that: The dual-loop optimization strategy of the active suspension system is specifically: Take the large loop iteration number j = 0, the small loop iteration number i = 0, and the change amount of each structural parameter of the actuator Changes in electro-magnetic-thermal coupling performance The quantitative influence degree is recorded as η. Based on η, the discrete interval z / η of the structural parameters of the Taguchi method experiment is obtained, and the sensitivity of each structural parameter of the actuator to the thrust fluctuation and the maximum temperature rise is calculated. Select sensitivity The structural parameters in the range of greater than 0.5 and less than -0.5 were fitted with response surface functions to establish the functional relationship between the highly sensitive structural parameters and the electromagnetic-thermal coupling performance; based on the functional relationship, the genetic algorithm was used to obtain the electro-magnetic-thermal coupling performance (T g ) j 、(F b ) j The Pareto frontier curve of the jth generation is obtained and recorded [w1, w2, w3...] j and the jth generation magnetothermal coupling performance (T g ) j 、(F b ) j , and judge whether the elements in the control coefficient combination set U are all taken. If the control coefficient combination set U is not taken, continue to cycle the above process, that is, j+1, until all the control coefficient combinations are taken, and compare the optimal values ​​of each generation. The smallest corresponding control parameter combination and actuator structure parameters are the global optimal parameter set, which are used as the design parameters of the active suspension system; in, z is the structural size ratio coefficient.

8. The electromagnetic-thermal dual-coupling temperature rise prediction and dual-cycle optimization method according to claim 7 is characterized in that: The process of obtaining the control coefficient combination set U is as follows: Based on the numerical analytical equation of the five-body two-degree-of-freedom active suspension, the ceiling damping constant coefficient C is obtained. sky , Passive damping constant coefficient C s The mapping relationship between the combination of and the root mean square value of suspension dynamic travel rms_s, the root mean square value of body acceleration rms_b, and the root mean square value of tire dynamic load rms_d is obtained, and then the constraint condition cloud diagram is obtained. In the constraint condition area that satisfies the electro-magnetic-thermal coupling performance optimization function, C sky , C s Take a set element at intervals to obtain a control parameter combination set U that meets the constraint conditions.

9. The electromagnetic-thermal dual-coupling temperature rise prediction and dual-cycle optimization method according to claim 8 is characterized in that: In order to obtain the change of each structural parameter of the actuator Changes in electro-magnetic-thermal coupling performance The quantitative influence of the actuator initial structural parameter combination W0 is obtained under the steady-state electro-magnetic-thermal field coupling characteristic set {I j , P j , T j }, continue to obtain the steady-state electric-magnetic-thermal field coupling characteristic set under other structural parameter combinations in the actuator structural parameter combination set W, and calculate different actuator structural parameter combinations. It is necessary to modify the force-electric coupling transfer unit in the electric-magnetic-thermal bidirectional coupling relationship transfer model of the active suspension system actuator. The numerical calculation equation in the transfer unit is: Where τ is the magnetic pole spacing of the actuator, L d and L q are the d-axis and q-axis inductances, respectively, and i d and i q are d and q axis currents respectively; Changes in actuator structural parameters will change τ, P ε , L d , L q , due to F d It is known that, therefore, i d and i q The value of changes, realizing the correction of the force-electric coupling transmission relationship.

10. The electromagnetic-thermal dual-coupling temperature rise prediction and dual-cycle optimization method according to claim 8, characterized in that: The numerical analytical equation of the five-body two-degree-of-freedom active suspension includes the comprehensive dynamic equation of the nonlinear five-body two-degree-of-freedom active suspension: Among them, Q j is the generalized external force, T is the overall kinetic energy of the five-body two-degree-of-freedom active suspension, q j is the generalized coordinate, V is the overall potential energy of the five-body two-degree-of-freedom active suspension, is the force on the ith particle, is the potential vector of the i-th particle, F ix is the component of the force on the ith particle on the x-axis, F iy is the component of the force on the ith particle on the y-axis, x i and i is the x and y axis coordinates of point i, and n is a constant; The overall kinetic energy of the five-body two-degree-of-freedom active suspension includes the kinetic energy of the tires. Among them, m u For tire quality, is the velocity of the hinge point A, is the angle change between the composite structure axis and the vertical direction; Moment of inertia of composite structure AB in five-body two-degree-of-freedom active suspension in, are the mass of composite component 2 in composite structure AB, the mass of composite component 1 in composite structure AB, the length of composite component 1, and the length of composite component 2, respectively. AN is the distance between the hinge point A and the center of mass N of the composite structure.

Images