CN117125138A - EPS angle control method and device based on single neuron algorithm - Google Patents

EPS angle control method and device based on single neuron algorithm Download PDF

Info

Publication number
CN117125138A
CN117125138A CN202311136978.6A CN202311136978A CN117125138A CN 117125138 A CN117125138 A CN 117125138A CN 202311136978 A CN202311136978 A CN 202311136978A CN 117125138 A CN117125138 A CN 117125138A
Authority
CN
China
Prior art keywords
eps
angle
single neuron
output torque
steering wheel
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Pending
Application number
CN202311136978.6A
Other languages
Chinese (zh)
Inventor
李支轶
杨佐智
彭文典
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Boshi Huayu Turning System Wuhan Co ltd
Original Assignee
Boshi Huayu Turning System Wuhan Co ltd
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Boshi Huayu Turning System Wuhan Co ltd filed Critical Boshi Huayu Turning System Wuhan Co ltd
Priority to CN202311136978.6A priority Critical patent/CN117125138A/en
Publication of CN117125138A publication Critical patent/CN117125138A/en
Pending legal-status Critical Current

Links

Classifications

    • B—PERFORMING OPERATIONS; TRANSPORTING
    • B62—LAND VEHICLES FOR TRAVELLING OTHERWISE THAN ON RAILS
    • B62D—MOTOR VEHICLES; TRAILERS
    • B62D6/00—Arrangements for automatically controlling steering depending on driving conditions sensed and responded to, e.g. control circuits
    • B—PERFORMING OPERATIONS; TRANSPORTING
    • B62—LAND VEHICLES FOR TRAVELLING OTHERWISE THAN ON RAILS
    • B62D—MOTOR VEHICLES; TRAILERS
    • B62D5/00—Power-assisted or power-driven steering
    • B62D5/04—Power-assisted or power-driven steering electrical, e.g. using an electric servo-motor connected to, or forming part of, the steering gear
    • B62D5/0457—Power-assisted or power-driven steering electrical, e.g. using an electric servo-motor connected to, or forming part of, the steering gear characterised by control features of the drive means as such
    • B62D5/046—Controlling the motor
    • G—PHYSICS
    • G05—CONTROLLING; REGULATING
    • G05B—CONTROL OR REGULATING SYSTEMS IN GENERAL; FUNCTIONAL ELEMENTS OF SUCH SYSTEMS; MONITORING OR TESTING ARRANGEMENTS FOR SUCH SYSTEMS OR ELEMENTS
    • G05B11/00—Automatic controllers
    • G05B11/01—Automatic controllers electric
    • G05B11/36—Automatic controllers electric with provision for obtaining particular characteristics, e.g. proportional, integral, differential
    • G05B11/42—Automatic controllers electric with provision for obtaining particular characteristics, e.g. proportional, integral, differential for obtaining a characteristic which is both proportional and time-dependent, e.g. P. I., P. I. D.
    • Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
    • Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
    • Y02T—CLIMATE CHANGE MITIGATION TECHNOLOGIES RELATED TO TRANSPORTATION
    • Y02T10/00—Road transport of goods or passengers
    • Y02T10/60—Other road transportation technologies with climate change mitigation effect
    • Y02T10/72—Electric energy management in electromobility

Landscapes

  • Engineering & Computer Science (AREA)
  • Chemical & Material Sciences (AREA)
  • Combustion & Propulsion (AREA)
  • Transportation (AREA)
  • Mechanical Engineering (AREA)
  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • Automation & Control Theory (AREA)
  • Steering Control In Accordance With Driving Conditions (AREA)

Abstract

The invention discloses an EPS angle control method based on a single neuron algorithm, which comprises the following steps: step S1, calculating an angle deviation value of a steering wheel; step S2, calculating the increment of the output torque of the booster motor according to a single neuron self-adaptive PID control strategy, wherein the single neuron self-adaptive PID control strategy adopts a supervised Hebb learning rule; step S3, adjusting the weighting coefficient through a supervised Hebb learning rule in the operation process; and S4, calculating to obtain an output torque value of the booster motor. According to the invention, under complex working conditions, the EPS control system can learn by self according to different working conditions by adopting the single neuron self-adaptive PID, so as to adapt to various working conditions, and solve the problem of poor applicability of common PID angle control under complex and changeable working conditions.

Description

EPS angle control method and device based on single neuron algorithm
Technical Field
The invention relates to the technical field of automobiles, in particular to an EPS angle control method and device based on a single neuron algorithm.
Background
In recent years, intelligent assisted driving and unmanned technologies are rapidly developing, and the implementation of these technologies is not separated from an Electric Power Steering (EPS), which is a very important device in the process of implementing steering control of an automobile.
Most EPS currently on the market use PID control strategies. In the actual running process, the PID control strategy obtains the current steering wheel target angle and the actual angle, calculates the difference value of the current steering wheel target angle and the actual angle, and controls the power-assisted motor to output torque. The parameters of the control strategy are usually fixed, and the control strategy has better control effect on paving with smaller gradient, but the control method with fixed parameters lacks the capability of adapting to environmental changes. With the progress of intelligent auxiliary driving and unmanned driving technologies, the driving working conditions of vehicles become more and more complex and various, and the requirement of accurate control cannot be met by adopting a PID control strategy with fixed parameters.
Disclosure of Invention
In order to solve the technical problems, the invention provides an EPS angle control method based on a single neuron algorithm,
the method comprises the following steps:
step S1, calculating an angle deviation value of a steering wheel;
step S2, calculating the increment of the output torque of the booster motor according to a single neuron self-adaptive PID control strategy, wherein the single neuron self-adaptive PID control strategy adopts a supervised Hebb learning rule;
step S3, adjusting the weighting coefficient through a supervised Hebb learning rule in the operation process;
and S4, calculating to obtain an output torque value of the booster motor.
The invention also provides an EPS angle control device based on the single neuron algorithm, which comprises a control module, a motor torque calculation module and an execution mechanism, wherein the control module judges whether an EPS angle control function can be activated according to the vehicle speed, the hand force of a driver, the steering wheel angle and the steering wheel rotating speed; the motor torque calculation module calculates the output torque of the power-assisted motor according to the EPS angle control activation state of the control module and the EPS angle control method based on the single neuron algorithm; the actuating mechanism is an EPS power-assisted motor and outputs torque according to the output torque value of the power-assisted motor.
The invention has the beneficial effects that:
under complex working conditions, the EPS control system can learn by itself according to different working conditions by adopting the single neuron self-adaptive PID, so as to adapt to various working conditions, and solve the problem of poor applicability of common PID angle control under complex and changeable working conditions.
Drawings
The invention is described in further detail below with reference to the attached drawings and detailed description:
FIG. 1 is a flow chart of a control method of the present invention
Detailed Description
Other advantages and technical effects of the present invention will become more fully apparent to those skilled in the art from the following disclosure, which is a detailed description of the present invention given by way of specific examples. The invention may be practiced or carried out in different embodiments, and details in this description may be applied from different points of view, without departing from the general inventive concept. It should be noted that the following embodiments and features in the embodiments may be combined with each other without conflict. The following exemplary embodiments of the present invention may be embodied in many different forms and should not be construed as limited to the specific embodiments set forth herein. It should be appreciated that these embodiments are provided so that this disclosure will be thorough and complete, and will fully convey the technical solution of these exemplary embodiments to those skilled in the art.
Example 1
FIG. 1 is a flow chart of a control method according to the embodiment, wherein Z -1 Is the control amount of the previous time.
The specific EPS angle control method based on the single neuron algorithm comprises the following steps:
step S1, calculating an angle deviation value of a steering wheel;
step S2, calculating the increment of the output torque of the booster motor according to a single neuron self-adaptive PID control strategy, wherein the single neuron self-adaptive PID control strategy adopts a supervised Hebb learning rule;
step S3, adjusting the weighting coefficient through a supervised Hebb learning rule in the operation process;
and S4, calculating to obtain an output torque value of the booster motor.
The specific implementation of the algorithm in each step is described in more detail below.
In step S1, the steering wheel angle signal is collected by an angle sensor, and the calculation formula of the angle deviation value of the steering wheel is as formula (1):
e＝ W cur -W tar (1)
wherein e is an angle deviation value, W tar For the target angle, W cur Is the current angle.
In the prior art, the expression of the conventional incremental PID control strategy is shown as the formula (2)
ΔU(k)＝K p (e(k)-e(k-1))+K i e(k)+K d (e(k)+e(k-2)-2e(k-1))
(2)
The median value DeltaU (K) is the increment of the output torque of the power-assisted motor at the moment K, e (K) is the difference value between the angle of the steering wheel at the moment K and the target angle, e (K-1) is the angle deviation value at the previous moment, e (K-2) is the angle deviation value at the next previous moment, and K p Is a proportionality coefficient, K i As integral coefficient, K d Is a differential coefficient.
In step S2, the specific method for calculating the increment of the output torque of the booster motor according to the single neuron adaptive PID control strategy is as follows:
introducing a single neuron algorithm, and improving a conventional incremental PID control strategy into a single neuron self-adaptive PID control strategy: using supervised Hebb learning rules, using o i Representing the activation value of a neuron i, using o j Representing the activation value, d, of another neuron j j To a desired output value, w ij Representing nervesThe link weights between element i and neuron j, where η is the learning rate, Δw ij (k) At time k w ij Change amount of o j (k) For the activation value of the neuron with another neuron j at time k, o i (k) The activation value of this neuron with another neuron i at time k. The expression of the supervised Hebb learning rule is as shown in formula (3):
Δw ij (k)＝η(d j (k)-o j (k))o j (k)o i (k) (3)
rewritable (2), record x 1 ＝e(k)，x 2 (k)＝e(k)-e(k-1)，x 3 (k) =e (k) -2e (k-1) +e (k-2), and the integral, proportional and differential coefficients are considered as x and x, respectively i (k) Corresponding weighting coefficient, the weighting coefficient is marked as omega' i (k) Adding a scaling factor K, and requiring K>0, the expression of the incremental PID can be rewritten as formula (4):
wherein the method comprises the steps of
The method for realizing self-adaption and self-learning capacity by the single neuron self-adaption PID control strategy is to continuously adjust the weighting coefficient in the running process. The adjustment of the weighting coefficient is realized through supervised Hebb learning, and the learning algorithm is obtained as formula (6):
Δω i (k)＝η i Z(k)U(k)x i (k) (6)
let eta 1 ＝η I ，η 2 ＝η P ，η 3 ＝η D ，η I 、η P And eta D Represents the integral, proportional and differential learning rates, respectively, since Z (k) is equivalent to d in equation (3) j (k)-o j (k) So at this time Z (k) =e (k), and U (k) =
U(k-1)+ΔU(k)，ω i (k)＝ω i (k-1)+Δω i (k) The calculation formula of the output torque value U (k) of the power-assisted motor in the k-moment single neuron self-adaptive PID control strategy can be obtained, namely, the control algorithm is as shown in formula (7):
connection strength ω at time k i (k) I.e. the learning algorithm is as in equation (8):
ω 1 (k)＝ω 1 (k-1)+η I Z(k)U(k)x 1 (k)
ω 2 (k)＝ω 2 (k-1)+η P Z(k)U(k)x 2 (k) (8)
ω 3 (k)＝ω 3 (k-1)+η D Z(k)U(k)x 3 (k)
and (3) synthesizing the formulas 1, 7 and 8 to obtain a control algorithm and a learning algorithm of the complete single neuron self-adaptive PID control strategy. Wherein U (k) is the output torque of the booster motor, and x 1 ＝e(k)，x 2 (k)＝e(k)-e(k-1)，x 3 (k) E (k) -2e (k-1) +e (k-2), where e (k) is the difference between the angle of the steering wheel at the current time and the target angle, e (k-1) is the angle deviation value at the previous time, and e (k-2) is the angle deviation value at the next previous time, η I 、η P And eta D Represents the integral, proportional and differential learning rates, respectively, Z (K) =e (K), K being the proportionality coefficient.
The single neuron self-adaptive PID control strategy adopts different learning rates eta for integral, proportion and differential coefficients respectively I 、η P η D When the software runs, the single neuron adaptive PID control strategy can be based on Z (k), U (k) and x at each moment i (k) Learning and adjusting omega in real time i (k) And the adjustment of the different weighting coefficients is realized.
Scaling factor K and learning rate eta P 、η I 、η D The value of (2) can be obtained by experiment and is preset in the EPS program.
Example 2
To implement the method of embodiment 1, this embodiment provides an EPS angle control device based on a single neuron algorithm, including: the control module, the motor torque calculation module and the actuating mechanism;
the control module judges whether the EPS angle control function can be activated according to conditions such as the vehicle speed, the hand force of a driver, the steering wheel angle, the steering wheel rotating speed and the like.
The motor torque calculation module controls the activation state, a preset coefficient K and a learning rate eta according to the EPS angle of the control module P 、η I 、η D Parameters the booster motor output torque was calculated as in example 1.
The executing mechanism is an EPS power-assisted motor and outputs torque according to the input target motor torque value of the calculating module.
The present invention has been described in detail by way of specific embodiments and examples, but these should not be construed as limiting the invention. Many variations and modifications may be made by one skilled in the art without departing from the principles of the invention, which is also considered to be within the scope of the invention.

Claims (9)

1. The EPS angle control method based on the single neuron algorithm is characterized by comprising the following steps of:
step S1, calculating an angle deviation value of a steering wheel;
step S2, calculating the increment of the output torque of the booster motor according to a single neuron self-adaptive PID control strategy, wherein the single neuron self-adaptive PID control strategy adopts a supervised Hebb learning rule;
step S3, adjusting the weighting coefficient through a supervised Hebb learning rule in the operation process;
and S4, calculating to obtain an output torque value of the booster motor.
2. The EPS angle control method based on the single neuron algorithm according to claim 1, characterized in that the angle deviation value of the steering wheel in step S1 is the current angle of the steering wheel minus the target angle.
3. The EPS angle control method based on the single neuron algorithm according to claim 1, characterized in that in the step S2, a calculation formula for calculating an increment of the output torque of the assist motor is as follows:
wherein DeltaU (K) is the increment of the output torque of the booster motor at the moment K, K is a proportionality coefficient omega' i (k) As a weighting coefficient, x i (k) The angle deviation value of the steering wheel at the previous moment is subtracted from the angle deviation value of the steering wheel at the moment k.
4. The EPS angle control method based on the single neuron algorithm according to claim 3, characterized in that the weighting coefficient ω' i (k) The calculation formula of (2) is as follows:
wherein omega i (k) The connection strength at time k.
5. The method for controlling the EPS angle based on the single neuron algorithm according to claim 4, characterized in that the connection strength ω at time k is i (k) The calculation formula of (2) is as follows:
ω 1 (k)＝ω 1 (k-1)+η I Z(k)U(k)x 1 (k)
ω 2 (k)＝ω 2 (k-1)+η P Z(k)U(k)x 2 (k)
ω 3 (k)＝ω 3 (k-1)+η D Z(k)U(k)x 3 (k)
wherein U (k) is the output torque eta of the booster motor I 、η P And eta D Represents the integral, proportional and differential learning rates, respectively, Z (k) =e (k), e (k) being the instant kThe angle deviation value of the steering wheel, K is a proportionality coefficient.
6. The EPS angle control method according to claim 5, characterized in that in the step S3, the calculation formula for adjusting the weighting coefficient is as follows:
Δω i (k)＝η i Z(k)U(k)x i (k)
wherein Δω i (k) The increment of the connection strength at time k.
7. The EPS angle control method based on the single-neuron algorithm according to claim 6, characterized in that in step S4, a calculation formula for calculating the output torque value of the assist motor is:
8. the method for controlling the angle of EPS based on the single neuron algorithm according to claim 7, characterized in that the proportional coefficient K, the integral, proportional and differential learning rate η P 、η I 、η D The value of (2) is obtained by a test and is set in advance in the EPS program.
9. The utility model provides an EPS angle control device based on single neuron algorithm, includes control module, motor torque calculation module and actuating mechanism, its characterized in that:
the control module judges whether an EPS angle control function can be activated according to the vehicle speed, the hand force of a driver, the steering wheel angle and the steering wheel rotating speed;
the motor torque calculation module calculates the output torque of the power-assisted motor according to the EPS angle control method based on the single neuron algorithm as set forth in any one of claims 1 to 8 according to the EPS angle control activation state of the control module;
the actuating mechanism is an EPS power-assisted motor and outputs torque according to the output torque value of the power-assisted motor.
CN202311136978.6A 2023-09-05 2023-09-05 EPS angle control method and device based on single neuron algorithm Pending CN117125138A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202311136978.6A CN117125138A (en) 2023-09-05 2023-09-05 EPS angle control method and device based on single neuron algorithm

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202311136978.6A CN117125138A (en) 2023-09-05 2023-09-05 EPS angle control method and device based on single neuron algorithm

Publications (1)

Publication Number Publication Date
CN117125138A true CN117125138A (en) 2023-11-28

Family

ID=88850715

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202311136978.6A Pending CN117125138A (en) 2023-09-05 2023-09-05 EPS angle control method and device based on single neuron algorithm

Country Status (1)

Country Link
CN (1) CN117125138A (en)

Similar Documents

Publication Publication Date Title
EP3459823B1 (en) Control device for electric power steering device
EP3459822B1 (en) Control device for electric power steering device
US9545949B2 (en) Steering control apparatus and a steering apparatus using the same
US20200010111A1 (en) Electric power steering apparatus
EP2050653B1 (en) Steering system
EP3495244B1 (en) Electric power steering apparatus
US10035538B2 (en) Electric power steering system with motor controller
JPWO2019107438A1 (en) Vehicle steering system
CN109911000B (en) Method for estimating steering wheel torque of vehicle
CN110182253A (en) The quadrant type friciton compensation that tyre load for steering system is estimated
CN108327781B (en) Electric steering type steering control device
CN111994167A (en) Electric power steering system
EP3939861B1 (en) Steering device
JPH058660A (en) Vehicle speed control device
JP2009280097A (en) Steering parameter optimization system
CN108197392B (en) Power-assisted characteristic curve design method of automobile electric power-assisted steering system based on SOC
CN116888025A (en) Control circuit for motor vehicle
CN117125138A (en) EPS angle control method and device based on single neuron algorithm
JP2002160653A (en) Control device of electric power steering device
WO2023100369A1 (en) Motor control device
CN105292243A (en) Prediction control method of electric power steering system of automobile
US8560177B2 (en) Electric power steering system
WO2023144895A1 (en) Motor control device
US20240059348A1 (en) Motor control in an electric power steering
WO2023062748A1 (en) Motor control device

Legal Events

Date Code Title Description
PB01 Publication
PB01 Publication
SE01 Entry into force of request for substantive examination
SE01 Entry into force of request for substantive examination