Hand force estimation method for steer-by-wire

Abstract

The invention relates to the technical field of automobile steer-by-wire, in particular to a hand power estimation method for steer-by-wire, which comprises the following steps: step 1, a signal preprocessing module processes a steering wheel angle signal, a steering wheel angular velocity signal and a motor torque signal; the method comprises the following steps: step 1, establishing a dynamical model of a steering wheel end and an extended state observer estimation method in an extended state observer module, and step 3, switching the estimated hand torque by a hand torque estimation switching module according to the output of the extended state observer module and outputting the hand torque to a downstream module. Compared with the prior art, the hand torque during normal driving is estimated by using the extended state observer and the small-range high-precision torque sensor motor torque, high precision of estimation is ensured, the hand torque under the tail end protection or get-on and get-off auxiliary working condition is estimated by using the motor torque, and under the condition that the hardware cost is not increased, the accuracy of estimation is ensured. And the demand of a steer-by-wire system on human hand force estimation is met.

Description

Technical Field

[0001] The invention relates to the technical field of automobile wire-controlled steering, in particular to a hand force estimation method for wire-controlled steering. Background Art

[0002] With the popularization of new energy vehicles and the rapid development of intelligent driving technology, the diversification and intelligence of automobile use needs have become an inevitable trend in the future development of automobiles. As a key system of intelligent driving, the steering system will also evolve from traditional mechanical steering, hydraulic steering, and electric power steering to wire-controlled steering to adapt to higher levels of autonomous driving and future intelligent transportation scenarios.

[0003] Steer-by-wire eliminates the mechanical intermediate shaft that connects the upper and lower parts of the traditional electric power steering system. Instead, it controls the steering actuator motor through electronic control signals, which drives the front wheels to complete the steering, thus achieving mechanical decoupling of the steering wheel and the steering front wheels of the car.

[0004] The wire-controlled steering system can better adapt to high-level assisted driving. When entering high-level assisted driving, the steering wheel can remain silent and be fully retracted into the smart cockpit through the electrically adjustable column to achieve true autonomous driving. Since the current laws and regulations have not been fully implemented, the current L2+-based assisted driving system still requires the participation of the driver. When high-level assisted driving is in operation, when danger is detected, the system needs to remind the driver to take over the vehicle manually. Or when the driver actively perceives that the current driving scene is dangerous, he will actively intervene in the control of the vehicle. At this time, the assisted driving system needs to identify the driver's intentions and exit the system, and transfer control of the vehicle to the driver.

[0005] At present, the mainstream method of judging the driver's steering intervention is to use a high-precision, small-range torque sensor. The torque sensor is installed on the steering column. When the steering system is working, the torque sensor is easily affected by the damping and friction from the steering column. Directly using the data of the torque sensor for judgment may cause the assisted driving system to exit by mistake due to interference, causing safety hazards to the driving vehicle.

[0006] Since the intermediate shaft is eliminated in the steer-by-wire system, the hand-feel simulation motor needs to generate a torque of 20-30Nm at the steering wheel end to prevent the driver from moving the steering wheel in the case of end-of-line protection or vehicle entry and exit assistance. At this time, the torque generated has far exceeded the range of traditional torque sensors.

[0007] If a high-precision and large-range torque sensor is used, the hardware cost will increase significantly. During normal driving, the driver's hand torque range is not very large, but the accuracy requirement will be higher. In the end protection or vehicle entry and exit assistance conditions, although the hand torque is large, the accuracy requirement is not high.

[0008] Therefore, it is necessary to design a hand force estimation method for wire-controlled steering to ensure high-precision estimation of the hand torque within the normal hand torque range and to output a reasonable estimated hand torque in real time under large hand torque conditions. Summary of the invention

[0009] The purpose of the present invention is to overcome the deficiencies of the prior art and provide a hand force estimation method for wire-controlled steering to ensure high-precision estimation of hand torque within a normal hand torque range and to output a reasonable estimated hand torque in real time under large hand torque conditions.

[0010] In order to achieve the above-mentioned purpose, the present invention is a hand force estimation method for wire-controlled steering, comprising the following steps: step 1, a signal preprocessing module processes a steering wheel angle signal, a steering wheel angular velocity signal, and a motor torque signal; step 2, a dynamic model of a steering wheel end and an expanded state observer estimation method are established in an extended state observer module; step 3, a hand torque estimation switching module switches the estimated hand torque according to the output of the expanded state observer module and outputs it to a downstream module.

[0011] The dynamic model of the steering wheel end is an equivalent system dynamic model of the steering wheel, torque sensor, steering column, and feel simulation motor. The steering wheel is connected to the end of the steering column, and the steering column is provided with a torque sensor and a feel simulation motor.

[0012] The said establishment of the dynamic model of the steering wheel end comprises the following steps: Step 1a, establishing the dynamic model of the steering wheel: Build the torque sensor model: T s =K s ·(θ s -θ c ), where J s is the steering wheel moment of inertia, B s is the steering wheel damping coefficient, θ s is the steering wheel angle, θ c is the steering column output angle, K s is the equivalent system friction force, T d Input hand torque for the steering wheel, T s The torque sensor measures the torque, g m is the transmission ratio of the motor reduction mechanism, r pg is the transmission ratio from rack to motor reduction mechanism, is the steering wheel angular speed, is the steering wheel angular acceleration; Step 2a, the dynamic model of the steering wheel and the torque sensor model are combined to obtain the torque balance formula for the upper steering: Step 3a, rewrite the moment balance differential equation into a state - space equation: Where, is the state variable, y is the output variable, θ s and are used as the system state variables, T s is used as the output variable of the state - space, θ c and T d are used as the known input variables of the state - space.

[0013] The estimation method of the extended state observer includes state expansion and reconstruction of the state - space equation, design of the extended state observer, and discretization of the extended state observer.

[0014] The state expansion and reconstruction of the state - space equation include the following steps: Step 1b, add the steering wheel input hand torque T d to the state variable to construct a new state variable Step 2b, the new state - space equation after expansion: Where, is the derivative of the expanded state variable, A new is the expanded system state matrix, B new is the expanded input matrix, C new is the expanded output matrix, C new =(K s 0 0).

[0015] The design of the extended state observer includes the following steps: Step 1c, complete the design of the extended state observer according to the space equation where, is the derivative of the state variable of the extended state observer, is the state variable of the extended state observer, is the output variable of the extended state observer, G is the gain matrix; Step 2c, assume the gain matrix G = [G 1 G 2 G 3 T , where, G1 is the gain coefficient 1, G2 is the gain coefficient 2, G3 is the gain coefficient 3, T is the transpose symbol, then the characteristic equation expression is f(s)=det[s·I-(A new -GC new )], where, s is the response of input signals with different frequencies, I is the identity matrix; Step 3c, assume the desired observer poles are a 1 、a 2 、a 3 ​, the characteristic equation expression is f(s) = (s + a 1 )(s + a 2 )(s + a 3 ); Step 4c, solve the two characteristic equations in Step 2c and Step 3c to obtain the specific parameter values of G1, G2, and G3; Step 4d, perform pole placement by building a simulation model to determine a reasonable value of the G matrix.

[0016] The discretization of the extended state observer includes the following steps: Step 1d, substitute the difference equation for : where T is the preset sampling time and k is the number of steps; Step 2d, discretize the designed extended state observer using the approximation method; Step 3d, after discretization, the current estimated state variable is derived from the input at the previous moment and the estimated state variable at the previous moment. Step 4d, perform phase lead processing on the estimated human hand torque.

[0017] The switching of the estimated hand torque is as follows: If the output of the extended state observer is greater than the torque threshold and satisfies the condition of the holding time threshold, then the estimated hand torque = the motor output torque × the worm and worm gear transmission ratio, and the direction is opposite to the motor torque; If the output of the extended state observer is less than or equal to the torque threshold or does not satisfy the condition of the holding time threshold, then the estimated hand torque = the output of the extended state observer.

[0018] The torque threshold is 10 Nm and the time threshold is 200 ms.

[0019] Compared with the prior art, the present invention uses an extended state observer and a torque sensor motor torque with a small range and high precision to estimate the hand torque during normal driving, ensuring high precision in estimation. It uses the motor torque to estimate the hand torque in the end protection or getting on and off the vehicle assistance conditions, meeting the requirements of the steer-by-wire system for human hand force estimation without increasing the hardware cost. The present invention has the advantages of high robustness, strong generalization, and rapid estimation. BRIEF DESCRIPTION OF THE DRAWINGS

[0020] Figure 1 is a schematic diagram of the dynamic model at the steering wheel end of the present invention.

[0021] Figure 2 is a flowchart of the present invention.

[0022] Figure 3 is a schematic diagram of the hand torque switching principle of the present invention.

[0023] Figure 4 is a diagram of the simulation results of the steering wheel angle tracking of the present invention.

[0024] Figure 5 It is a tracking simulation result diagram of the hand torque and the input hand torque of the present invention. DETAILED DESCRIPTION

[0025] The present invention will now be further described with reference to the accompanying drawings.

[0026] See also Figure 2 The present invention provides a hand force estimation method for wire-controlled steering, comprising the following steps: Step 1, a signal preprocessing module processes a steering wheel angle signal, a steering wheel angular velocity signal, and a motor torque signal.

[0027] Step 2: Establish the dynamic model of the steering wheel end and the extended state observer estimation method in the extended state observer module.

[0028] The extended state observer is designed based on the state-space equation, which is derived from the equivalent dynamics model and expands the hand torque to be estimated into a new state quantity.

[0029] See also Figure 1 The dynamic model of the steering wheel end is an equivalent system dynamic model of the steering wheel 1, the torque sensor 2, the steering column 3, and the feel simulation motor 4. The steering wheel 1 is connected to the end of the steering column 3, and the steering column 3 is provided with a torque sensor 2 and a feel simulation motor 4.

[0030] Establishing the dynamic model of the steering wheel includes the following steps: Step 1a, establishing the dynamic model of the steering wheel: Build the torque sensor model: T s =K s ·(θ s -θ c ), where J s is the steering wheel moment of inertia, B s is the steering wheel damping coefficient, θ s is the steering wheel angle, θ c is the steering column output angle, K s is the equivalent system friction force, T d Input hand torque for the steering wheel, T s The torque sensor measures the torque, g m is the transmission ratio of the motor reduction mechanism, r pg is the transmission ratio from rack to motor reduction mechanism, is the steering wheel angular speed, is the steering wheel angular acceleration; Step 2a, the dynamic model of the steering wheel and the torque sensor model are combined to obtain the torque balance formula for the upper steering: Step 3a, rewrite the torque balance differential equation into a state space equation: Among them, is the state variable, y is the output variable, and θ s and are used as system state variables, and T s is used as the output variable of the state space, and θ c and T d are used as the known input variables of the state space.

[0031] The extended state observer estimation method includes state expansion and reconstruction of the state space equation, design of the extended state observer, and discretization of the extended state observer.

[0032] State expansion and reconstruction of the state space equation include the following steps: Step 1b, add the steering wheel input hand torque T d to the state variable to construct a new state variable Step 2b, the new state space equation after expansion: Among them, is the derivative of the expanded state variable, A new is the expanded system state matrix, B new is the expanded input matrix, C new is the expanded output matrix, and C new =(K s 0 0).

[0033] Design of the extended state observer includes the following steps: Step 1c, complete the design of the extended state observer according to the space equation , among which, is the derivative of the state variable of the extended state observer, is the state variable of the extended state observer, is the output variable of the extended state observer, and G is the gain matrix; Step 2c, assume the gain matrix G = [G 1 G 2 G 3 , T , among which, G1 is the gain coefficient 1, G2 is the gain coefficient 2, G3 is the gain coefficient 3, and T is the transpose symbol. Then the characteristic equation expression is f(s) = det[s·I-(A new -GC new )], where s is the response of the input signal at different frequencies, and I is the identity matrix; Step 3c, assume the desired observer poles are a 1 , a 2 , a 3 , then the characteristic equation expression is f(s) = (s + a 1 )(s + a2 )(s + a 3 ); Step 4c, solve the two characteristic equations in Step 2c and Step 3c to obtain the specific parameter values of G1, G2, and G3; Step 4d, perform pole placement by building a simulation model to determine a reasonable value of the G matrix.

[0034] The discretization of the extended state observer includes the following steps: Step 1d, substitute for the difference equation: where T is the preset sampling time and k is the number of step lengths; Step 2d, discretize the designed extended state observer using an approximation method; Step 3d, after discretization, the current estimated state variable is derived from the input at the previous moment and the estimated state variable at the previous moment, that is, the current hand torque can be calculated from the hand torque value estimated at the previous moment and the rotation angle of the output end of the steering column as the input.

[0035] Step 3, the hand torque estimation switching module switches the estimated hand torque according to the output of the extended state observer module and outputs it to the downstream module.

[0036] In the cases of getting in and out of the vehicle assistance or end protection, the hand torque will be much greater than the range of the torque sensor. At this time, the upper limit of the hand torque that the extended disturbance observer can estimate will be limited by the range of the torque sensor, and the true hand torque cannot be estimated. Since the upper steering friction is small, on the premise of ignoring the friction. The motor torque acting on the steering wheel end can be regarded as entirely overcoming the hand torque of the person. At this time, the magnitude of the hand torque of the person is approximately equal to the motor torque at the steering wheel end and is opposite to the direction of the motor torque.

[0037] Refer to Figure 3 , and switch the estimated hand torque as follows: If the output of the extended state observer is greater than the torque threshold and satisfies the condition of the hold time threshold, then the estimated hand torque = the motor output torque × the worm gear transmission ratio, and the direction is opposite to the motor torque; If the output of the extended state observer is less than or equal to the torque threshold or does not satisfy the condition of the hold time threshold, then the estimated hand torque = the output of the extended state observer. Preferably, the torque threshold is 10 Nm and the time threshold is 200 ms.

[0038] The present invention uses the extended state observer algorithm to estimate the human hand force in real time. The steering wheel angle tracking simulation result diagram and the tracking simulation result diagram of the hand torque and the input hand torque of the present invention are as shown in Figure 4 , Figure 5As shown, the present invention uses an extended state observer and a torque sensor motor torque with a small range and high precision to estimate the hand torque magnitude during normal driving, ensuring high-precision estimation. The motor torque is used to estimate the hand torque magnitude under the conditions of end protection or getting on and off the vehicle assistance. Without increasing the hardware cost, the requirements of the steer-by-wire system for human hand force estimation are met. The present invention has the advantages of high robustness, strong generalization, and rapid estimation.

Claims

1. A hand force estimation method for wire-controlled steering, characterized in that: The method comprises the following steps: step 1, a signal preprocessing module processes a steering wheel angle signal, a steering wheel angular velocity signal and a motor torque signal; step 2, a dynamic model of a steering wheel end and an extended state observer estimation method are established in an extended state observer module; step 3, a hand torque estimation switching module switches the estimated hand torque according to the output of the extended state observer module and outputs it to a downstream module.

2. The hand force estimation method for wire-controlled steering according to claim 1, characterized in that: The dynamic model of the steering wheel end is an equivalent system dynamic model of a steering wheel (1), a torque sensor (2), a steering column (3), and a hand-feel simulation motor (4); the steering wheel (1) is connected to the end of the steering column (3), and the steering column (3) is provided with a torque sensor (2) and a hand-feel simulation motor (4).

3. The hand force estimation method for wire-controlled steering according to claim 1, characterized in that: The said establishment of the dynamic model of the steering wheel end comprises the following steps: Step 1a, establishing the dynamic model of the steering wheel: Build the torque sensor model: T s =K s ·(θ s -θ c ), where Js is the steering wheel moment of inertia, B s is the steering wheel damping coefficient, θ s is the steering wheel angle, θ c is the steering column output angle, K s is the equivalent system friction force, T d Input hand torque for the steering wheel, T s The torque sensor measures the torque, g m is the transmission ratio of the motor reduction mechanism, r pg is the transmission ratio from rack to motor reduction mechanism, is the steering wheel angular speed, is the steering wheel angular acceleration; Step 2a, the dynamic model of the steering wheel and the torque sensor model are combined to obtain the torque balance formula for the upper steering: Step 3a, rewrite the torque balance differential equation into a state space equation: in, is the state quantity, y is the output quantity, θ s and As the system state variable, T s As the output quantity in the state space, θ c and T d As known input quantities in the state space.

4. The hand force estimation method for wire-controlled steering according to claim 1, characterized in that: The extended state observer estimation method includes state expansion and reconstruction of state space equations, extended state observer design, and discretization of the extended state observer.

5. The hand force estimation method for wire-controlled steering according to claim 4, characterized in that: The state expansion and reconstruction of the state space equations include the following steps: Step 1b, inputting the steering wheel hand torque T d Add to state In the process, we construct a new state variable Step 2b, the new state space equation after expansion: in, To find the derivative of the expanded state, A new is the expanded system state matrix, B new is the expanded input matrix, C new is the expanded output matrix, C new =(K s 0 0).

6. The hand force estimation method for wire-controlled steering according to claim 4, characterized in that: The extended state observer design comprises the following steps: Step 1c, according to the space equation Complete the design of the extended state observer, where is the derivative of the state quantity of the extended state observer, is the state variable of the extended state observer, is the output of the extended state observer, G is the gain matrix; Step 2c, let the gain matrix G = [G1 G2 G3] T , where G1 is gain coefficient 1, G2 is gain coefficient 2, G3 is gain coefficient 3, and T is the transposition symbol. The characteristic equation is f(s) = det[s·I-(A new -GC new ), where s is the response of the input signal of different frequencies, and I is the unit matrix; step 3c, assuming that the desired observer poles are a1, a2, and a3, the characteristic equation expression is f(s) = (s+a1)(s+a2)(s+a3); step 4c, solve the two characteristic equations in step 2c and step 3c to obtain the specific parameter values ​​of G1, G2, and G3; step 4d, perform pole configuration by building a simulation model to determine the reasonable value of the G matrix.

7. The hand force estimation method for wire-controlled steering according to claim 4, characterized in that: The discretization of the extended state observer includes the following steps: Step 1d, Make a substitution in the difference equation: Where T is the preset sampling time, k is the number of steps; Step 2d, discretize the designed extended state observer using the approximation method; Step 3d, after discretization, the current estimated state variable The input at the last moment and the state variable estimated at the last moment It is deduced that, in step 4d, phase advance processing is performed on the estimated hand torque.

8. The hand force estimation method for wire-controlled steering according to claim 1, characterized in that: The switching of the estimated hand torque is as follows: if the output of the expanded state observer is greater than the torque threshold and meets the condition of the holding time threshold, the estimated hand torque = motor output torque × worm gear transmission ratio, and the direction is opposite to the motor torque; if the output of the expanded state observer is less than or equal to the torque threshold or does not meet the condition of the holding time threshold, the estimated hand torque = the output of the expanded state observer.

9. The hand force estimation method for wire-controlled steering according to claim 8, characterized in that: The torque threshold is 10 Nm, and the time threshold is 200 ms.

Images