CN109455186A - Three-wheel independently drives the hybrid optimization of electri forklift torque to distribute control method - Google Patents

Abstract

The invention discloses a kind of three-wheels, and the hybrid optimization of electri forklift torque independently to be driven to distribute control method, the ideal yaw velocity and side slip angle computation model of electri forklift are independently driven using three-wheel, are calculated and are obtained ideal yaw velocity and ideal side slip angle that three-wheel independently drives electri forklift；According to the difference between the true yaw velocity and ideal yaw velocity in fork truck driving process, and the difference between real centroid side drift angle and ideal side slip angle, it is calculated using genetic algorithm cooperation fork truck six degrees of freedom model and obtains antero posterior axis distribution coefficient and Y-axis distribution coefficient；The driving moment for obtaining three wheels is calculated according to total driving moment, antero posterior axis distribution coefficient and Y-axis distribution coefficient, the yaw velocity and side slip angle that the method for the present invention makes fork truck are close to ideal value, so that three-wheel be made independently electri forklift to be driven to have better stability and flexibility.

Description

Hybrid optimization distribution control method for torque of three-wheel independent drive electric forklift

Technical Field

The invention relates to the field of safe auxiliary driving and intelligent control, in particular to a stability control method for a three-wheel independently-driven electric forklift.

Background

The four-wheel independent drive electric vehicle is a research hotspot in the field of pure electric vehicles at home and abroad at present. Four drive wheel torques of the four-wheel independent drive electric automobile can be distributed at will, the control is flexible, the stability control of the electric forklift can be realized by distributing the four drive wheel torques, and the method is a new idea for controlling the stability of the automobile.

The forklift is important equipment in a warehouse in the logistics industry, and the traditional forklift has high energy consumption, serious environmental pollution and poor stability; for this reason, three-wheeled independently-driven electric forklifts are popular. Due to the difference of the structure and the working environment, the control method of the four-wheel independent drive electric vehicle in the prior art cannot be directly applied to the three-wheel independent drive electric forklift, and the direct application of the control method can greatly reduce the stability and flexibility of the three-wheel independent drive electric forklift.

Disclosure of Invention

The invention aims to overcome the defects in the prior art, and provides a hybrid optimal distribution control method for the torque of the three-wheel independently-driven electric forklift, so that the three-wheel independently-driven electric forklift with advantages in energy consumption and environmental pollution effectively ensures the stability of the three-wheel independently-driven electric forklift in a complex working environment through optimal control.

The invention adopts the following technical scheme for solving the technical problems:

the invention relates to a hybrid optimal distribution control method for torque of a three-wheel independent driving electric forklift, wherein three independent driving wheels in the three-wheel independent driving electric forklift are respectively a left front wheel, a right front wheel and a rear wheel, the rear wheel is a steering wheel, and the optimal distribution control method is characterized in that:

the method comprises the steps that an ideal yaw angular velocity calculation model and an ideal mass center slip angle calculation model of the three-wheel independent driving electric forklift are established, and the ideal yaw angular velocity and the ideal mass center slip angle of the three-wheel independent driving electric forklift are calculated and obtained;

calculating by utilizing a genetic algorithm and a three-wheel independent drive forklift six-degree-of-freedom model according to the difference between the real yaw angular velocity and the ideal yaw angular velocity in the running process of the forklift and the difference between the real centroid slip angle and the ideal centroid slip angle to obtain a front-rear axis distribution coefficient and a left-right axis distribution coefficient;

and calculating to obtain the driving torques of the three wheels according to the total driving torque, the front-rear shaft distribution coefficient and the left-right shaft distribution coefficient, and driving each wheel in a one-to-one correspondence manner to enable the yaw angular velocity and the mass center slip angle of the forklift to be close to ideal values.

The invention discloses a hybrid optimized distribution control method of three-wheel independent drive electric forklift torque, which is characterized in that: the optimized distribution control method comprises the following steps:

step 1: determining the ideal yaw velocity omega of the forklift in the current driving state according to the current driving state of the forklift_rdAnd ideal centroid slip angle β_rd；

Step 2: detecting and obtaining actual yaw velocity omega of forklift in current driving state_realAnd actual centroid slip angle β_realThen there is the actual difference value omega of the yaw rate_gAnd the actual difference value β of centroid slip angle_gRespectively as follows: omega_g＝ω_real-ω_rd，β_g＝β_real-β_rd；

If the actual difference value omega of the yaw angular velocity_gAnd the actual difference value β of centroid slip angle_gAre all less than 0.2, indicating an actual yaw rate ω_realAnd actual centroid slip angle β_realAll meet the stability requirement, and the final value k of the distribution coefficient of the front and rear axes_frzAnd final value k of left-right axis distribution coefficient_lrzAll are set to 0.5, and step 5 is continuously executed, otherwise step 3 is continuously executed;

and step 3: according to the actual difference value omega of the yaw angular velocity_gAnd the actual difference value β of centroid slip angle_gCalculating by adopting a genetic algorithm to obtain the current value k of the distribution coefficient of the front axis and the rear axis_frcAnd the current value k of left-right axis distribution coefficient_lrc；

And 4, step 4: according to the current value k of the distribution coefficient of the front and rear axes_frcLeft and right axis distribution coefficient current value k_lrcAnd total torque T of driving wheel of forklift_{General assembly}Calculating to obtain current values of torques of three driving wheels of the forklift; obtaining through calculation by utilizing a six-degree-of-freedom model of three-wheel independent driving electric forklift according to current values of torques of three driving wheels of forkliftSimulating yaw angular velocity omega_vAnd simulated centroid slip angle β_vThen there is a yaw rate analog difference value omega_rvSimulated difference from centroid slip angle β_rvRespectively as follows: omega_rv＝ω_v-ω_rd，β_rv＝β_v-β_rd；

If the yaw angular velocity simulates the difference value omega_rvSimulated difference from centroid slip angle β_rvAre all less than 0.2, indicating a simulated yaw rate ω_vAnd simulated centroid slip angle β_vAll meet the stability requirement, and the final value k of the distribution coefficient of the front and rear axes_frzAnd final value k of left-right axis distribution coefficient_lrzOne-to-one correspondence is set as the current value k of the distribution coefficient of the front and rear axes_frcLeft and right axis distribution coefficient current value k_lrcAnd continuing to execute the step 5; otherwise, updating the population in the genetic algorithm and returning to the step 3;

and 5: using the front and rear axes distribution coefficient final value k_frzLeft and right axis distribution coefficient final value k_lrzAnd total torque T of driving wheel of forklift_{General assembly}Calculating to obtain the final torque value T of three driving wheels of the forklift_d1z、T_d2z、T_d3zAnd each driving wheel of the forklift is independently controlled in a one-to-one correspondence manner;

step 6: and (5) circularly executing the steps 1 to 5 at set time intervals to realize the stability control of the three-wheel independent driving electric forklift.

The method for controlling the torque mixing optimization distribution of the three-wheel independently-driven electric forklift is also characterized in that in the step 1, the ideal mass center slip angle β of the forklift in the current driving state is used_rdSet to 0, i.e., β_rdWhen the yaw rate is 0, the ideal yaw rate omega of the forklift in the current driving state is determined as follows_rd：

Longitudinal running speed v of forklift in current running state is obtained by utilizing sensor measurement_xSteering wheel angle delta and lateral travel speed v_y(ii) a Calculating and obtaining the ideal yaw angular velocity omega of the forklift by the formula (1)_rd：

Wherein,k is the stability factor of the three-wheeled forklift,mu is the road surface friction coefficient, C₁For front wheel cornering stiffness, C₃The lateral deflection rigidity of the rear wheel is shown, a is the distance from the center of mass of the forklift to the front shaft, b is the distance from the center of mass of the forklift to the rear shaft, g is the gravity acceleration, m is the total mass of the forklift, and L is a + b.

The invention discloses a hybrid optimized distribution control method of three-wheel independent drive electric forklift torque, which is characterized in that: in the step 4, a current value k is distributed according to the front and rear axes_frcAnd the current value k of left-right axis distribution coefficient_lrcAnd total torque T of driving wheel of forklift_{General assembly}The current values of the torques of three driving wheels in the forklift are obtained by calculation according to the formula (2), wherein the current values are respectively as follows: current value of left front wheel input torque T_d1cCurrent value of right front wheel input torque T_d2cAnd the current value T of the input torque of the rear wheel_d3c；

The invention discloses a hybrid optimized distribution control method of three-wheel independent drive electric forklift torque, which is characterized in that: in the step 5, the final values of the torques of the three driving wheels in the forklift are obtained by calculating according to the formula (3), wherein the final values are respectively: left front wheel input torque final value T_d1zRight front wheel input torque final value T_d2zAnd a final value T of the input torque of the rear wheel_d3z：

The invention discloses a hybrid optimized distribution control method of three-wheel independent drive electric forklift torque, which is characterized in that: a six-degree-of-freedom model of the three-wheel independent driving electric forklift is established as follows:

determining a longitudinal force equation of the forklift as formula (4):

determining a lateral force equation of the forklift as shown in formula (5):

F_x1for left front wheel longitudinal force, F_x2Is the longitudinal force of the right front wheel, F_x3For rear wheel longitudinal forces, F_y1For left front wheel lateral forces, F_y2Is a right front wheel lateral force, F_y3A transverse force of the rear wheels, a_yFor lateral acceleration, a_xIn the form of a longitudinal acceleration, the acceleration,is v is_xThe amount of the differential of (a) is,is v is_yThe differential amount of (a);

determining a forklift yaw torque balance equation of the three-wheel independently-driven electric forklift as shown in the formula (6):

wherein c is the center distance between the left front wheel and the right front wheel of the forklift, β is the centroid slip angle, I_zFor rotation of the whole vehicle about the z-axisThe inertia moment of the rotor is generated,is omega_rDifferential of, ω_rIs the yaw rate of the vehicle;

the wheel dynamics equation of the three-wheel independently driven electric forklift characterized by equation (7) is established according to the moment balance principle:

I_z1is the rotational inertia of the left front wheel, I_z2Is the moment of inertia of the right front wheel, I_z3For rear wheel moment of inertia, T_d1Is the left front wheel torque, T_d2Is the right front wheel torque, T_d3As rear wheel torque, omega₁Angular velocity of rotation of the left front wheel, ω₂Is the angular velocity, ω, of the right front wheel rotation₃Rotational angular velocity of rear wheel, R_wIs the wheel rolling radius;

establishing a vertical load calculation equation of the three driving wheels according to the formula (8):

F_z1for vertical loading of the left front wheel, F_z2For right front wheel vertical load, F_z3The vertical load of the rear wheel is adopted, and h is the height of the center of mass of the forklift;

determining the slip angle equations of the three driving wheels, which are respectively:

left front wheel side slip angle sigma₁Comprises the following steps:

right front wheel slip angle sigma₂Comprises the following steps:

rear wheel side slip angle sigma₃Comprises the following steps:

tire models for three wheels were determined: the universal tire model is a tire model of a three-wheel independent driving electric forklift;

calculating to obtain the longitudinal speed of each driving wheel, wherein the longitudinal speed is respectively as follows:

left front wheel longitudinal speed v₁Comprises the following steps:

front right wheel longitudinal velocity v₂Comprises the following steps:

rear wheel longitudinal speed v₃Comprises the following steps: v. of₃＝v_x+(v_y+aω_r)sinδ；

And calculating to obtain the wheel slip ratio of each drive, wherein the wheel slip ratio is respectively as follows:

left front wheel slip ratio s₁Comprises the following steps:

slip ratio s of front right wheel₂Comprises the following steps:

rear wheel slip ratio s₃Comprises the following steps:

the invention relates to a three-wheel independently-driven electric forklift torqueThe hybrid optimal allocation control method is also characterized in that: in the step 3, an optimized genetic algorithm is adopted, and the current value k of the distribution coefficient of the front and rear axes is obtained by calculation according to the following process_frcAnd the current value k of left-right axis distribution coefficient_lrc：

Step 3.1 randomly generating N binary numbers with 8 bits and respectively recording the N binary numbers as individuals α_iAs an initial population Q₁，i＝1,2...N；

Step 3.2 calculation to obtain α Each Individual_iThe current value k of the corresponding front and rear axis distribution coefficient_frciAnd the current value k of left-right axis distribution coefficient_lrciIs to make the individual α_iIs converted into a 10-system number and is used as k_frciOf (2) will be α_iIs converted into a 10-system number and is used as k_lrci；

Step 3.3 calculation to obtain α Each Individual_iFitness function value J of_iComprises the following steps:

in the formula, ω_ωr、ω_β、ω_frAnd ω_lrIs the set weight value;

defining optimal individuals α_bestThe individual with the minimum fitness value;

step 3.4: selecting M individuals from N individuals as an update group Q by adopting a roulette method in a genetic algorithm for selection operation, wherein the roulette method is a selection strategy based on fitness proportion₂Go to step 3.5, M is less than N, α for each individual_iProbability of being selected p_iComprises the following steps:

step 3.5: adopting the law of universal gravitation to carry out the cross operation of the genetic algorithm:

step 3.5.1: obtaining an updated population Q by calculation of equation (9)₂α of each individual_jPassive gravitational mass M_pjAnd active gravitational mass M_aj，j＝1,2...M：

Wherein, J_iIs an individual α_jFitness function value of, maxJ_iAnd minJ_iRespectively being individual α_jMaximum and minimum fitness of m_jIs an individual α_jInertial mass of, M_jIs an individual α_jThe gravitational mass of;

step 3.5.2: calculating to obtain updated population Q₂Force between any two individuals

Definition of individuals α_pTo individual α_qActing force F_pqComprises the following steps:

wherein p ≠ q ≠ M, 2.. M, q ≠ 1,2.. M;

M_ppis an individual α_pMass of passive attraction, M_aqIs an individual α_qActive gravitational mass of;

g is a constant of universal gravitation, G ═ G₀e^-T，G₀The initial value of the universal gravitation constant is set, and T is the iteration number of the genetic algorithm;

R_pqis an individual α_pAnd α_qThe distance between the two or more of the two or more,

3.5.3 obtaining individuals α by calculation using the formula (10)_pSubject α_qPost-acting acceleration A_pq：

Wherein M is_pIs an individual α_pThe gravitational mass of;

3.5.4 Each individual α_jα corresponding to population optimal individual_bestPerforming crossover operation to obtain α each individual in one-to-one correspondence_jα of the updated individual_j' from all updated individuals α_j' formation update population Q₃Updating individual α_j' obtained by calculation of equation (11):

α_j′＝(1-A_jbest)α_best+A_jbestα_j(11)，

wherein A is_jbestIs an individual α_jSubject to population optimization individual α_best(ii) post-applied acceleration;

step 3.6: mutation operation of genetic algorithms using adaptive mutation strategies

Step 3.6.1: the updated group Q is obtained by calculation from the equation (12) respectively₃α of each update individual_j' adaptive mutation probability p_mj：

p_mReference value, p, representing the probability of adaptive mutation_minLower value limit, p, representing the probability of adaptive mutation_maxUpper value limit, p, representing the probability of adaptive mutation_m、p_minAnd p_maxIs manually selected to be between 0 and 1, and: p is a radical of_min＜p_m＜p_max；

J_i' update individual α_j' fitness function value of, J_max' to update the population Q₃Maximum value of fitness of all individuals in (1), J_min' to update the population Q₃The minimum value of the fitness of all individuals in the group,representing an update group Q₃Average value of fitness of all individuals in the population;

step 3.6.2: using the adaptive mutation probability p_mjCompleting the variation operation of each updated individual, and obtaining the final individuals α in a one-to-one correspondence manner_j″；

Step 3.7 calculation of all Final individuals α_j"and finding the final individual α with the least fitness function value_best", if α_best"has a fitness function value less than or equal to 0.5, then the individual α is identified_best"the high four bits are converted into 10-system numbers and used as k_frcOf (2) will be α_best"the lower four bits are converted into 10-ary numbers and used as k_lrcIf α_best"has a fitness function value greater than 0.5, then all of the final individuals α are utilized_jAnd generating a new generation of population by adding a plurality of newly added and randomly generated binary number individuals with 8 bits, and returning the new generation of population to the step 3.2 for next iteration, wherein the number of the individuals of the new generation of population is N.

Compared with the prior art, the invention has the beneficial effects that:

1. according to the structure and the working environment of the three-wheel independent drive electric forklift, a six-degree-of-freedom whole vehicle model of the three-wheel independent drive electric forklift and an ideal yaw angular velocity calculation model of the three-wheel independent drive electric forklift are established; the universal gravitation law and the adaptive variation strategy are adopted to optimize the genetic algorithm, and the optimized genetic algorithm is utilized to reasonably distribute the torques of the three driving wheels of the three-wheel independent driving electric forklift, so that the three-wheel independent driving electric forklift has better stability and flexibility under various complex working conditions.

2. The optimized genetic algorithm has higher calculation efficiency and calculation precision, can avoid the genetic algorithm from falling into local optimization to a certain extent, and can reasonably distribute the driving wheel torque of the three-wheel independent driving electric forklift so that the three-wheel independent driving electric forklift has good stability under various complex working conditions.

Drawings

FIG. 1 is a flow chart of a control method of the present invention;

FIG. 2 is a six-degree-of-freedom model of a three-wheel independently driven electric forklift in the invention;

Detailed Description

Referring to fig. 1 and fig. 2, in the embodiment, three independent driving wheels in the three-wheel independent drive electric forklift are a left front wheel, a right front wheel and a rear wheel, respectively, the rear wheel is a steering wheel, and the hybrid optimal distribution control method of the three-wheel independent drive electric forklift torque is performed as follows:

the ideal yaw angular velocity and the ideal barycenter slip angle of the three-wheel independently driven electric forklift are calculated and obtained by establishing an ideal yaw angular velocity calculation model and an ideal barycenter slip angle calculation model of the three-wheel independently driven electric forklift.

According to the difference value between the real yaw angular velocity and the ideal yaw angular velocity in the running process of the forklift and the difference value between the real centroid slip angle and the ideal centroid slip angle, a front-rear axis distribution coefficient and a left-right axis distribution coefficient are obtained through calculation by utilizing a genetic algorithm and matching with a three-wheel independent drive forklift six-degree-of-freedom model.

And calculating to obtain the driving torques of the three wheels according to the total driving torque, the front-rear shaft distribution coefficient and the left-right shaft distribution coefficient, and driving each wheel in a one-to-one correspondence manner to enable the yaw angular velocity and the mass center slip angle of the forklift to be close to ideal values.

In this embodiment, the hybrid optimal distribution control method for the torques of the three-wheel independently-driven electric forklift is performed according to the following steps:

step 1: determining the ideal yaw velocity omega of the forklift in the current driving state according to the current driving state of the forklift_rdAnd ideal centroid slip angle β_rdTo obtain a reference value in control.

Step 2: detecting and obtaining actual yaw velocity omega of forklift in current driving state_realAnd actual centroid slip angle β_realThen there is the actual difference value omega of the yaw rate_gAnd the actual difference value β of centroid slip angle_gRespectively as follows: omega_g＝ω_real-ω_rd，β_g＝β_real-β_rd；

If the actual difference value omega of the yaw angular velocity_gAnd the actual difference value β of centroid slip angle_gAre all less than 0.2, indicating an actual yaw rate ω_realAnd actual centroid slip angle β_realAll meet the stability requirement, and the final value k of the distribution coefficient of the front and rear axes_frzAnd final value k of left-right axis distribution coefficient_lrzAll set to 0.5, the total torque is evenly distributed to the three drive wheels and step 5 is continued, otherwise step 3 is continued.

And step 3: according to the actual difference value omega of the yaw angular velocity_gAnd the actual difference value β of centroid slip angle_gCalculating by adopting a genetic algorithm to obtain the current value k of the distribution coefficient of the front axis and the rear axis_frcAnd the current value k of left-right axis distribution coefficient_lrcBecause the genetic algorithm has the rapid random search capability, the current value k of the distribution coefficient of the front axle and the rear axle which can stabilize the forklift can be rapidly and accurately obtained_frcAnd the current value k of left-right axis distribution coefficient_lrc。

And 4, step 4: in order to ensure that the front-rear axis distribution coefficient obtained in step 3 is currentValue k_frcAnd the current value k of left-right axis distribution coefficient_lrcThe wheel torque can be reasonably distributed to enable the running state of the vehicle to tend to be stable, and the current value k of the distribution coefficient of the front axle and the rear axle needs to be adjusted_frcAnd the current value k of left-right axis distribution coefficient_lrcInputting a six-degree-of-freedom model of the three-wheel independent driving electric forklift to simulate the running state of the forklift. According to the current value k of the distribution coefficient of the front and rear axes_frcLeft and right axis distribution coefficient current value k_lrcAnd total torque T of driving wheel of forklift_{General assembly}Calculating to obtain current values of torques of three driving wheels of the forklift; obtaining the simulated yaw angular velocity omega by utilizing the six-degree-of-freedom model calculation of the three-wheel independent driving electric forklift according to the current values of the torques of the three driving wheels of the forklift_vAnd simulated centroid slip angle β_vThen there is a yaw rate analog difference value omega_rvSimulated difference from centroid slip angle β_rvRespectively as follows: omega_rv＝ω_v-ω_rd，β_rv＝β_v-β_rd；

If the yaw angular velocity simulates the difference value omega_rvSimulated difference from centroid slip angle β_rvAre all less than 0.2, indicating a simulated yaw rate ω_vAnd simulated centroid slip angle β_vAll meet the stability requirement, and the front and rear axis distribution coefficient current value k obtained in the step 3_frcAnd the current value k of left-right axis distribution coefficient_lrcThe wheel torque can be reasonably distributed, the running state of the vehicle tends to be stable, and the final value k of the front and rear axle distribution coefficient is used_frzAnd final value k of left-right axis distribution coefficient_lrzOne-to-one correspondence is set as the current value k of the distribution coefficient of the front and rear axes_frcLeft and right axis distribution coefficient current value k_lrcAnd continuing to execute the step 5; otherwise, updating the population in the genetic algorithm and returning to the step 3.

And 5: using the front and rear axes distribution coefficient final value k_frzLeft and right axis distribution coefficient final value k_lrzAnd total torque T of driving wheel of forklift_{General assembly}Calculating to obtain the final torque value T of three driving wheels of the forklift_d1z、T_d2z、T_d3zAnd each driving wheel of the forklift is independently controlled in one-to-one correspondence。

Step 6: and (5) circularly executing the steps 1 to 5 at set time intervals to realize the stability control of the three-wheel independent driving electric forklift.

In the specific implementation, the corresponding measures also comprise:

in step 1, the ideal mass center slip angle β under the current driving state of the forklift is determined_rdSet to 0, i.e., β_rdWhen the yaw rate is 0, the ideal yaw rate omega of the forklift in the current driving state is determined as follows_rd：

Longitudinal running speed v of forklift in current running state is obtained by utilizing sensor measurement_xSteering wheel angle delta and lateral travel speed v_y(ii) a Calculating and obtaining the ideal yaw angular velocity omega of the forklift by the formula (1)_rd：

Wherein,k is the stability factor of the three-wheeled forklift,mu is the friction coefficient of the road surface, the road surface of the forklift is smoother, therefore, the friction coefficient mu of the road surface is 0.8, C₁The front wheel yaw stiffness is a front wheel yaw stiffness having a value close to that of the front wheel yaw stiffness, and is collectively denoted by C₁，C₃The lateral deflection rigidity of the rear wheel is shown, a is the distance from the center of mass of the forklift to the front shaft, b is the distance from the center of mass of the forklift to the rear shaft, g is the gravity acceleration, m is the total mass of the forklift, and L is a + b.

In step 4, the current value k is assigned according to the front and rear axes_frcAnd the current value k of left-right axis distribution coefficient_lrcAnd forklift truck driveTotal torque T of driving wheel_{General assembly}The current values of the torques of three driving wheels in the forklift are obtained by calculation according to the formula (2), wherein the current values are respectively as follows: current value of left front wheel input torque T_d1cCurrent value of right front wheel input torque T_d2cAnd the current value T of the input torque of the rear wheel_d3c；

In step 5, the final torque values of the three driving wheels in the forklift are obtained by calculating according to the formula (3), wherein the final torque values are the final input torque values T of the left front wheel_d1zRight front wheel input torque final value T_d2zAnd a final value T of the input torque of the rear wheel_d3z：

In specific implementation, a six-degree-of-freedom model of the three-wheel independent driving electric forklift is established as follows:

determining a longitudinal force equation of the forklift as formula (4):

determining a lateral force equation of the forklift as shown in formula (5):

F_x1for left front wheel longitudinal force, F_x2Is the longitudinal force of the right front wheel, F_x3For rear wheel longitudinal forces, F_y1For left front wheel lateral forces, F_y2Is a right front wheel lateral force, F_y3A transverse force of the rear wheels, a_yFor lateral acceleration, a_xIs longitudinal acceleration，Is v is_xThe amount of the differential of (a) is,is v is_yThe differential amount of (a);

determining a forklift yaw torque balance equation of the three-wheel independently-driven electric forklift as shown in the formula (6):

wherein c is the center distance between the left front wheel and the right front wheel of the forklift, β is the centroid slip angle, I_zIs the rotational inertia of the whole vehicle around the z-axis,is omega_rDifferential of, ω_rIs the yaw rate of the vehicle;

the wheel dynamics equation of the three-wheel independently driven electric forklift characterized by equation (7) is established according to the moment balance principle:

I_z1is the rotational inertia of the left front wheel, I_z2Is the moment of inertia of the right front wheel, I_z3For rear wheel moment of inertia, T_d1Is the left front wheel torque, T_d2Is the right front wheel torque, T_d3As rear wheel torque, omega₁Angular velocity of rotation of the left front wheel, ω₂Is the angular velocity, ω, of the right front wheel rotation₃Rotational angular velocity of rear wheel, R_wIs the wheel rolling radius, the wheel rolling radius R_wSelecting according to the vehicle model;

establishing a vertical load calculation equation of the three driving wheels according to the formula (8):

F_z1for vertical loading of the left front wheel, F_z2For right front wheel vertical load, F_z3The vertical load of the rear wheel is adopted, and h is the height of the center of mass of the forklift;

determining the slip angle equations of the three driving wheels, which are respectively:

left front wheel side slip angle sigma₁Comprises the following steps:

right front wheel slip angle sigma₂Comprises the following steps:

rear wheel side slip angle sigma₃Comprises the following steps:

tire models for three wheels were determined: the method comprises the following steps of adopting a universal tire model as a tire model of a three-wheel independent driving electric forklift, wherein at present, the universal tire model is a magic formula model, a power index unified tire model or a SWIFT tire model;

calculating to obtain the longitudinal speed of each driving wheel, wherein the longitudinal speed is respectively as follows:

left front wheel longitudinal speed v₁Comprises the following steps:

front right wheel longitudinal velocity v₂Comprises the following steps:

rear wheel longitudinal speed v₃Comprises the following steps: v. of₃＝v_x+(v_y+aω_r)sinδ；

And calculating to obtain the wheel slip ratio of each drive, wherein the wheel slip ratio is respectively as follows:

left front wheel slip ratio s₁Comprises the following steps:

slip ratio s of front right wheel₂Comprises the following steps:

rear wheel slip ratio s₃Comprises the following steps:

in step 3, an optimized genetic algorithm is adopted to calculate and obtain the current value k of the distribution coefficient of the front axis and the rear axis according to the following process_frcAnd the current value k of left-right axis distribution coefficient_lrc：

Step 3.1 randomly generating N binary numbers with 8 bits and respectively recording the N binary numbers as individuals α_iAs an initial population Q₁，i＝1,2...N；

Step 3.2 calculation to obtain α Each Individual_iThe current value k of the corresponding front and rear axis distribution coefficient_frciAnd the current value k of left-right axis distribution coefficient_lrciIs to make the individual α_iIs converted into a 10-system number and is used as k_frciOf (2) will be α_iIs converted into a 10-system number and is used as k_lrciFor example, α_i00010010, then α will be_iThe high four bits of (A) are converted into a 10-system as k_frci：k_frci1, will α_iThe lower four bits of (A) are converted into a 10-system as k_lrci：k_lrci＝2。

Step 3.3 calculation to obtain α Each Individual_iFitness function value J of_iComprises the following steps:

in the formula, ω_ωr、ω_β、ω_frAnd ω_lrIs the set weight value;

defining optimal individuals α_bestThe individual with the minimum fitness value;

step 3.4: selecting by roulette method based on fitness proportion, and selecting M individuals as update group Q₂Go to step 3.5, M is less than N, α for each individual_iProbability of being selected p_iComprises the following steps:

step 3.5: and (4) performing genetic algorithm cross operation by adopting the law of universal gravitation. By using the universal gravitation search algorithm and the local search thought for reference, the cross operation in the genetic algorithm is improved, the universal gravitation search algorithm is introduced into the cross operation of the genetic algorithm, the optimal individual information can be fully utilized, the convergence speed is accelerated, and the solving precision is improved:

step 3.5.1: obtaining an updated population Q by calculation of equation (9)₂α of each individual_jPassive gravitational mass M_pjAnd active gravitational mass M_aj，j＝1,2...M：

Wherein, J_iIs an individual α_jFitness function value of, maxJ_iAnd minJ_iAre respectively an individualα_jMaximum and minimum fitness of m_jIs an individual α_jInertial mass of, M_jIs an individual α_jThe gravitational mass of;

step 3.5.2: calculating to obtain updated population Q₂Force between any two individuals

Definition of individuals α_pTo individual α_qActing force F_pqComprises the following steps:

wherein p ≠ q ≠ M, 2.. M, q ≠ 1,2.. M;

M_ppis an individual α_pMass of passive attraction, M_aqIs an individual α_qActive gravitational mass of;

g is a constant of universal gravitation, G ═ G₀e^-T，G₀Is an initial value of a constant of universal gravitation, G₀Taking 1.25 and T as the iteration number of the genetic algorithm;

R_pqis an individual α_pAnd α_qThe distance between the two or more of the two or more,

3.5.3 obtaining individuals α by calculation using the formula (10)_pSubject α_qPost-acting acceleration A_pq：

Wherein M is_pIs an individual α_pThe gravitational mass of;

3.5.4 Each individual α_jα corresponding to population optimal individual_bestTo carry out intersectionFork operation, one-to-one obtaining each individual α_jα of the updated individual_j' from all updated individuals α_j' formation update population Q₃Updating individual α_j' obtained by calculation of equation (11):

α_j′＝(1-A_jbest)α_best+A_jbestα_j(11)，

wherein A is_jbestIs an individual α_jSubject to population optimization individual α_best(ii) post-applied acceleration;

step 3.6: and carrying out mutation operation of the genetic algorithm by using an adaptive mutation strategy. The self-adaptive mutation strategy is used for carrying out mutation operation of the genetic algorithm, so that the predicament that the genetic algorithm is trapped in local optimum can be effectively avoided, and the convergence speed is greatly accelerated.

Step 3.6.1: the updated group Q is obtained by calculation from the equation (12) respectively₃α of each update individual_j' adaptive mutation probability p_mj：

p_mReference value, p, representing the probability of adaptive mutation_minLower value limit, p, representing the probability of adaptive mutation_maxUpper value limit, p, representing the probability of adaptive mutation_m、p_minAnd p_maxIs manually selected to be between 0 and 1, and: p is a radical of_min＜p_m＜p_max；

J_i' update individual α_j' fitness function value of, J_max' to update the population Q₃Maximum value of fitness of all individuals in (1), J_min' to update the population Q₃The minimum value of fitness of all individuals in the population, J represents the updated population Q₃Average value of fitness of all individuals in the population;

step 3.6.2: use the instituteThe adaptive mutation probability p_mjCompleting the variation operation of each updated individual, and obtaining the final individuals α in a one-to-one correspondence manner_j″；

Step 3.7 calculation of all Final individuals α_j"and finding the final individual α with the least fitness function value_best", if α_best"has a fitness function value less than or equal to 0.5, then the individual α is identified_best"the high four bits are converted into 10-system numbers and used as k_frcOf (2) will be α_best"the lower four bits are converted into 10-ary numbers and used as k_lrcIf α_best"has a fitness function value greater than 0.5, then all of the final individuals α are utilized_jAnd generating a new generation of population by adding a plurality of newly added and randomly generated binary number individuals with 8 bits, and returning the new generation of population to the step 3.2 for next iteration, wherein the number of the individuals of the new generation of population is N.

Claims (7)

1. A hybrid optimal distribution control method for torque of a three-wheel independent drive electric forklift is characterized in that the optimal distribution control method comprises the following steps:

the method comprises the steps that an ideal yaw angular velocity calculation model and an ideal mass center slip angle calculation model of the three-wheel independent driving electric forklift are established, and the ideal yaw angular velocity and the ideal mass center slip angle of the three-wheel independent driving electric forklift are calculated and obtained;

calculating by utilizing a genetic algorithm and a three-wheel independent drive forklift six-degree-of-freedom model according to the difference between the real yaw angular velocity and the ideal yaw angular velocity in the running process of the forklift and the difference between the real centroid slip angle and the ideal centroid slip angle to obtain a front-rear axis distribution coefficient and a left-right axis distribution coefficient;

and calculating to obtain the driving torques of the three wheels according to the total driving torque, the front-rear shaft distribution coefficient and the left-right shaft distribution coefficient, and driving each wheel in a one-to-one correspondence manner to enable the yaw angular velocity and the mass center slip angle of the forklift to be close to ideal values.

2. The hybrid optimal distribution control method for the torques of the three-wheeled independently driven electric forklifts according to claim 1, wherein the optimal distribution control method is performed by the following steps:

step 1: determining the ideal yaw velocity omega of the forklift in the current driving state according to the current driving state of the forklift_rdAnd ideal centroid slip angle β_rd；

Step 2: detecting and obtaining actual yaw velocity omega of forklift in current driving state_realAnd actual centroid slip angle β_realThen there is the actual difference value omega of the yaw rate_gAnd the actual difference value β of centroid slip angle_gRespectively as follows: omega_g＝ω_real-ω_rd，β_g＝β_real-β_rd；

If the actual difference value omega of the yaw angular velocity_gAnd the actual difference value β of centroid slip angle_gAre all less than 0.2, indicating an actual yaw rate ω_realAnd actual centroid slip angle β_realAll meet the stability requirement, and the final value k of the distribution coefficient of the front and rear axes_frzAnd final value k of left-right axis distribution coefficient_lrzAll are set to 0.5, and step 5 is continuously executed, otherwise step 3 is continuously executed;

and step 3: according to the actual difference value omega of the yaw angular velocity_gAnd the actual difference value β of centroid slip angle_gCalculating by adopting a genetic algorithm to obtain the current value k of the distribution coefficient of the front axis and the rear axis_frcAnd the left and right axesCurrent value k of coefficient_lrc；

And 4, step 4: according to the current value k of the distribution coefficient of the front and rear axes_frcLeft and right axis distribution coefficient current value k_lrcAnd total torque T of driving wheel of forklift_{General assembly}Calculating to obtain current values of torques of three driving wheels of the forklift; obtaining the simulated yaw angular velocity omega by utilizing the six-degree-of-freedom model calculation of the three-wheel independent driving electric forklift according to the current values of the torques of the three driving wheels of the forklift_vAnd simulated centroid slip angle β_vThen there is a yaw rate analog difference value omega_rvSimulated difference from centroid slip angle β_rvRespectively as follows: omega_rv＝ω_v-ω_rd，β_rv＝β_v-β_rd；

If the yaw angular velocity simulates the difference value omega_rvSimulated difference from centroid slip angle β_rvAre all less than 0.2, indicating a simulated yaw rate ω_vAnd simulated centroid slip angle β_vAll meet the stability requirement, and the final value k of the distribution coefficient of the front and rear axes_frzAnd final value k of left-right axis distribution coefficient_lrzOne-to-one correspondence is set as the current value k of the distribution coefficient of the front and rear axes_frcLeft and right axis distribution coefficient current value k_lrcAnd continuing to execute the step 5; otherwise, updating the population in the genetic algorithm and returning to the step 3;

and 5: using the front and rear axes distribution coefficient final value k_frzLeft and right axis distribution coefficient final value k_lrzAnd total torque T of driving wheel of forklift_{General assembly}Calculating to obtain the final torque value T of three driving wheels of the forklift_d1z、T_d2z、T_d3zAnd each driving wheel of the forklift is independently controlled in a one-to-one correspondence manner;

step 6: and (5) circularly executing the steps 1 to 5 at set time intervals to realize the stability control of the three-wheel independent driving electric forklift.

3. The method as claimed in claim 2, wherein the step 1 comprises determining the ideal centroid slip angle β under the current driving condition of the forklift_rdIs set to 0β is_rdWhen the yaw rate is 0, the ideal yaw rate omega of the forklift in the current driving state is determined as follows_rd：

Longitudinal running speed v of forklift in current running state is obtained by utilizing sensor measurement_xSteering wheel angle delta and lateral travel speed v_y(ii) a Calculating and obtaining the ideal yaw angular velocity omega of the forklift by the formula (1)_rd：

Wherein,k is the stability factor of the three-wheeled forklift,mu is the road surface friction coefficient, C₁For front wheel cornering stiffness, C₃The lateral deflection rigidity of the rear wheel is shown, a is the distance from the center of mass of the forklift to the front shaft, b is the distance from the center of mass of the forklift to the rear shaft, g is the gravity acceleration, m is the total mass of the forklift, and L is a + b.

4. The hybrid optimal distribution control method of three-wheeled individual drive electric forklift torque according to claim 2, characterized in that: in the step 4, a current value k is distributed according to the front and rear axes_frcAnd the current value k of left-right axis distribution coefficient_lrcAnd total torque T of driving wheel of forklift_{General assembly}The current values of the torques of three driving wheels in the forklift are obtained by calculation according to the formula (2), wherein the current values are respectively as follows: current value of left front wheel input torque T_d1cCurrent value of right front wheel input torque T_d2cAnd the current value T of the input torque of the rear wheel_d3c；

5. According toThe hybrid optimal distribution control method of three-wheeled individual drive electric forklift torque according to claim 2, characterized in that: in the step 5, the final values of the torques of the three driving wheels in the forklift are obtained by calculating according to the formula (3), wherein the final values are respectively: left front wheel input torque final value T_d1zRight front wheel input torque final value T_d2zAnd a final value T of the input torque of the rear wheel_d3z：

6. The hybrid optimal distribution control method of three-wheeled individual drive electric forklift torque according to claim 2, characterized in that: a six-degree-of-freedom model of the three-wheel independent driving electric forklift is established as follows:

determining a longitudinal force equation of the forklift as formula (4):

determining a lateral force equation of the forklift as shown in formula (5):

F_x1for left front wheel longitudinal force, F_x2Is the longitudinal force of the right front wheel, F_x3For rear wheel longitudinal forces, F_y1For left front wheel lateral forces, F_y2Is a right front wheel lateral force, F_y3A transverse force of the rear wheels, a_yFor lateral acceleration, a_xIn the form of a longitudinal acceleration, the acceleration,is v is_xThe amount of the differential of (a) is,is v is_yThe differential amount of (a);

determining a forklift yaw torque balance equation of the three-wheel independently-driven electric forklift as shown in the formula (6):

wherein c is the center distance between the left front wheel and the right front wheel of the forklift, β is the centroid slip angle, I_zIs the rotational inertia of the whole vehicle around the z-axis,is omega_rDifferential of, ω_rIs the yaw rate of the vehicle;

establishing a wheel dynamic equation of the three-wheel independent drive electric forklift characterized by the formula (7) according to a moment balance principle:

I_z1is the rotational inertia of the left front wheel, I_z2Is the moment of inertia of the right front wheel, I_z3For rear wheel moment of inertia, T_d1Is the left front wheel torque, T_d2Is the right front wheel torque, T_d3As rear wheel torque, omega₁Angular velocity of rotation of the left front wheel, ω₂Is the angular velocity, ω, of the right front wheel rotation₃Rotational angular velocity of rear wheel, R_wIs the wheel rolling radius;

establishing a vertical load calculation equation of the three driving wheels according to the formula (8):

F_z1for vertical loading of the left front wheel, F_z2For right front wheel vertical load, F_z3The vertical load of the rear wheel is adopted, and h is the height of the center of mass of the forklift;

determining the slip angle equations of the three driving wheels, which are respectively:

left front wheel side slip angle sigma₁Comprises the following steps:

right front wheel slip angle sigma₂Comprises the following steps:

rear wheel side slip angle sigma₃Comprises the following steps:

tire models for three wheels were determined: the universal tire model is a tire model of a three-wheel independent driving electric forklift;

calculating to obtain the longitudinal speed of each driving wheel, wherein the longitudinal speed is respectively as follows:

left front wheel longitudinal speed v₁Comprises the following steps:

front right wheel longitudinal velocity v₂Comprises the following steps:

rear wheel longitudinal speed v₃Comprises the following steps: v. of₃＝v_x+(v_y+aω_r)sinδ；

And calculating to obtain the wheel slip ratio of each drive, wherein the wheel slip ratio is respectively as follows:

left front wheel slip ratio s₁Comprises the following steps:

slip ratio s of front right wheel₂Comprises the following steps:

rear wheel slip ratio s₃Comprises the following steps:

7. the hybrid optimal distribution control method for the torques of the three-wheeled individual drive electric forklift as recited in claim 2, wherein in said step 3, the current value k of the distribution coefficient of the front and rear axles is obtained by calculation using an optimal genetic algorithm according to the following process_frcAnd the current value k of left-right axis distribution coefficient_lrc：

Step 3.1 randomly generating N binary numbers with 8 bits and respectively recording the N binary numbers as individuals α_iAs an initial population Q₁，i＝1,2...N；

Step 3.2 calculation to obtain α Each Individual_iThe current value k of the corresponding front and rear axis distribution coefficient_frciAnd the current value k of left-right axis distribution coefficient_lrciIs to make the individual α_iIs converted into a 10-system number and is used as k_frciOf (2) will be α_iIs converted into a 10-system number and is used as k_lrci；

Step 3.3 calculation to obtain α Each Individual_iFitness function value J of_iComprises the following steps:

in the formula, ω_ωr、ω_β、ω_frAnd ω_lrIs the set weight value;

defining optimal individuals α_bestThe individual with the minimum fitness value;

step 3.4: selecting M individuals from N individuals as an update group Q by adopting a roulette method in a genetic algorithm for selection operation, wherein the roulette method is a selection strategy based on fitness proportion₂Go to step 3.5, M is less than N, α for each individual_iProbability of being selected p_iComprises the following steps:

step 3.5: adopting the law of universal gravitation to carry out the cross operation of the genetic algorithm:

step (ii) of3.5.1: obtaining an updated population Q by calculation of equation (9)₂α of each individual_jPassive gravitational mass M_pjAnd active gravitational mass M_aj，j＝1,2...M：

Wherein, J_iIs an individual α_jFitness function value of, maxJ_iAnd minJ_iRespectively being individual α_jMaximum and minimum fitness of m_jIs an individual α_jInertial mass of, M_jIs an individual α_jThe gravitational mass of;

step 3.5.2: calculating to obtain updated population Q₂Force between any two individuals

Definition of individuals α_pTo individual α_qActing force F_pqComprises the following steps:

wherein p ≠ q ≠ M, 2.. M, q ≠ 1,2.. M;

M_ppis an individual α_pMass of passive attraction, M_aqIs an individual α_qActive gravitational mass of;

g is a constant of universal gravitation, G ═ G₀e^-T，G₀The initial value of the universal gravitation constant is set, and T is the iteration number of the genetic algorithm;

R_pqis an individual α_pAnd α_qThe distance between the two or more of the two or more,

3.5.3 obtaining individuals α by calculation using the formula (10)_pSubject α_qPost-acting acceleration A_pq：

Wherein M is_pIs an individual α_pThe gravitational mass of;

3.5.4 Each individual α_jα corresponding to population optimal individual_bestPerforming crossover operation to obtain α each individual in one-to-one correspondence_jα of the updated individual_j' from all updated individuals α_j' formation update population Q₃Updating individual α_j' obtained by calculation of equation (11):

α_j′＝(1-A_jbest)α_best+A_jbestα_j(11)，

wherein A is_jbestIs an individual α_jSubject to population optimization individual α_best(ii) post-applied acceleration;

step 3.6: mutation operation of genetic algorithms using adaptive mutation strategies

Step 3.6.1: the updated group Q is obtained by calculation from the equation (12) respectively₃α of each update individual_j' adaptive mutation probability p_mj：

p_mReference value, p, representing the probability of adaptive mutation_minLower value limit, p, representing the probability of adaptive mutation_maxUpper value limit, p, representing the probability of adaptive mutation_m、p_minAnd p_maxIs manually selected to be between 0 and 1, and: p is a radical of_min＜p_m＜p_max；

J_i' update individual α_j' fitness function value of, J_max' to update the population Q₃Maximum value of fitness of all individuals in (1), J_min' to update the population Q₃The minimum value of the fitness of all individuals in the group,representing an update group Q₃Average value of fitness of all individuals in the population;

step 3.6.2: using the adaptive mutation probability p_mjCompleting the variation operation of each updated individual, and obtaining the final individuals α in a one-to-one correspondence manner_j″；

Step 3.7 calculation of all Final individuals α_j"and finding the final individual α with the least fitness function value_best", if α_best"has a fitness function value less than or equal to 0.5, then the individual α is identified_best"the high four bits are converted into 10-system numbers and used as k_frcOf (2) will be α_best"the lower four bits are converted into 10-ary numbers and used as k_lrcIf α_best"has a fitness function value greater than 0.5, then all of the final individuals α are utilized_jAnd generating a new generation of population by adding a plurality of newly added and randomly generated binary number individuals with 8 bits, and returning the new generation of population to the step 3.2 for next iteration, wherein the number of the individuals of the new generation of population is N.