JP7148058B2 - vehicle steering system - Google Patents

Description

The present invention relates to a vehicle steering system.

2. Description of the Related Art There is a vehicle steering apparatus capable of changing a steering angle ratio, which is a ratio between a steering angle of a steering wheel and a steering angle of a steered wheel. The steering angle ratio is also called an overall ratio of steering, an overall gear ratio, a steering gear ratio, or the like. Patent Literature 1 discloses a variable steering angle ratio steering device that changes a steering angle ratio (a ratio in which the denominator is the amount of change in the steering angle and the numerator is the amount of change in the steering angle) based on the vehicle speed of the vehicle. This variable steering angle ratio steering device controls the steering angle ratio to be small when the vehicle speed is high and to be large when the vehicle speed is low.

JP-A-5-105106

In a vehicle steering system capable of changing the steering angle ratio, for example, in a low speed range of a vehicle, by changing the steering angle ratio so that the amount of change in the turning angle with respect to the amount of change in the steering angle becomes large, Control may be performed to reduce the amount of steering by the driver. In this state, it is possible to steer with good reactivity to the driver's steering. However, when the vehicle turns at a very low speed, such as when turning at an intersection, the steering operation that is sensitive to the driver's operation may destabilize the running of the vehicle. Therefore, it may become difficult for the driver to drive.

SUMMARY OF THE INVENTION Accordingly, the present invention provides a vehicle steering system that reduces the influence of changes in the steering angle ratio on running.

A vehicle steering system according to an aspect of the present invention is a vehicle steering system that independently controls an output steering angle output to a steering mechanism with respect to an input steering angle input from a steering wheel. a steering angle calculator for determining a ratio of the input steering angle to the target steering angle and calculating the target steering angle according to the input steering angle and the ratio; correcting the target steering angle; and a turning angle correction unit that outputs the output turning angle, and the turning angle correction unit corrects the target turning angle so that the smaller the ratio, the greater the suppression of the high frequency component of the target turning angle.

According to the vehicle steering system of the present invention, it is possible to reduce the influence of changes in the steering angle ratio on driving.

FIG. 1 is a diagram schematically showing an example of the overall configuration of a vehicle steering system according to Embodiment 1. FIG. FIG. 2 is a diagram showing an example of a functional configuration of the ECU of FIG. 1; 3A is a diagram showing the relationship between the frequency of a target yaw rate signal and the gain according to Embodiment 1. FIG. 3B is a diagram showing the relationship between the frequency of the target yaw rate signal and the phase delay according to Embodiment 1. FIG. FIG. 4 is a diagram showing an example of a functional configuration of an ECU in a vehicle steering system according to Embodiment 2. As shown in FIG. FIG. 5 is a diagram showing an example of a configuration for learning a filter model according to Embodiment 2. FIG.

A vehicle steering system according to an embodiment will be described below with reference to the drawings. It should be noted that the embodiments described below represent comprehensive or specific examples. Numerical values, shapes, materials, constituent elements, arrangement positions of constituent elements, connection modes, and the like shown in the following embodiments are examples, and are not intended to limit the present invention. In addition, among the constituent elements in the following embodiments, constituent elements that are not described in independent claims representing the highest concept will be described as arbitrary constituent elements. Each figure is a schematic diagram and is not necessarily strictly illustrated. Furthermore, in each drawing, substantially the same components are denoted by the same reference numerals, and redundant description may be omitted or simplified.

[Embodiment 1]
A configuration of a vehicle steering system 1 according to Embodiment 1 of the present invention will be described. In the present embodiment, the vehicle steering system 1 is described as constituting a steer-by-wire system in which the steering mechanism 2 and the steering mechanism 3 are separated without being mechanically connected. FIG. 1 schematically shows an example of the overall configuration of a vehicle steering system 1 according to Embodiment 1. As shown in FIG. A vehicle steering system 1 is mounted on a vehicle A. As shown in FIG. As shown in FIG. 1, the vehicle steering system 1 includes a steering mechanism 2 operated by a driver of a vehicle A, and a steering mechanism that steers steerable wheels 80 according to an input to the steering mechanism 2 by the driver. 3. The steered wheels 80 are wheels for steering the vehicle A. FIG. The steering mechanism 2 and the steering mechanism 3 are separated without being mechanically connected.

The steering mechanism 2 includes a steering wheel 10 as a steering member operated by the driver for steering, a steering shaft 20 connected to the steering wheel 10, a reaction force motor 31 that applies a reaction force to the steering shaft 20, and a first reduction gear 40 that transmits the rotational driving force of the reaction motor 31 to the steering shaft 20 . The steering mechanism 2 includes a steering angle sensor 50 on the steering shaft 20 for detecting the amount of rotation of the steering shaft 20 , that is, the steering angle of the steering wheel 10 .

The first reduction gear 40 is connected to the steering shaft 20 and the reaction motor 31 , and applies a reaction force to the steering to the steering shaft 20 using the reaction motor 31 as a drive source. As a result, when the driver operates the steering wheel 10, the driver can feel the weight of the steering wheel 10 and the feeling of steering. The first speed reducer 40 reduces the rotational speed of the reaction motor 31 and increases the rotational driving force for transmission to the steering shaft 20 . For example, the first speed reducer 40 is composed of at least one gear that is in gear engagement with the steering shaft 20 and the driving rotation shaft of the reaction motor 31 .

The vehicle steering system 1 also includes an ECU (Electronic Control Unit) 100 . The ECU 100 is electrically connected to the reaction force motor 31 and a steering motor 32 of the steering mechanism 3 to be described later, and controls the operations of the reaction force motor 31 and the steering motor 32 . The steered motor 32 steers the steered wheels 80 by its rotational driving force. The ECU 100 is also electrically connected to a speed sensor 91 of the vehicle A, a battery 92, a steering angle sensor 50, and a steering angle sensor 71, which will be described later. The ECU 100 acquires a vehicle speed signal from the speed sensor 91 , a steering angle signal from the steering angle sensor 50 , and a steering angle signal of the steered wheels 80 from the steering angle sensor 71 . The ECU 100 supplies the electric power of the battery 92 to the reaction force motor 31 and the steering motor 32 to operate them. The ECU 100 controls the torque generated by the reaction force motor 31 according to the steering angle of the steering wheel 10 detected by the steering angle sensor 50 and the vehicle speed detected by the speed sensor 91 . The ECU 100 determines the turning angle of the steerable wheels 80 (also called “target turning angle”) based on information obtained from the steering angle sensor 50 , the speed sensor 91 and the turning angle sensor 71 . The ECU 100 controls the amount of rotation and torque of the steering motor 32 to operate the steered wheels 80 so that the steered angle of the steered wheels 80 becomes equal to the target steered angle. Communication among the ECU 100, the reaction motor 31, the steering motor 32, the speed sensor 91, the battery 92, the steering angle sensor 50, the steering angle sensor 71, etc. is communication via an in-vehicle network such as CAN (Controller Area Network). may be

The ECU 100 may be configured by a microcomputer having a processor such as a CPU (Central Processing Unit) or a DSP (Digital Signal Processor) and a memory. Examples of memory may be volatile memory such as RAM (Random Access Memory) and non-volatile memory such as ROM (Read-Only Memory). A part or all of the functions of the ECU 100 may be achieved by the CPU executing a program recorded in the ROM using the RAM as a working memory.

The steering mechanism 3 includes a rack shaft 70 connected to the steered wheels 80, a steering motor 32, a second reduction gear 60 that transmits the rotational driving force of the steering motor 32 to the rack shaft 70, and a steering angle sensor. 71. The steered angle sensor 71 is a sensor for detecting the steered angle of the steered wheels 80 . The steering angle sensor 71 is provided on the driving rotary shaft of the steering motor 32 . The ECU 100 converts the rotation angle of the drive rotary shaft detected by the steering angle sensor 71 into the rotation angle of the steered direction of the steered wheels 80 and determines the steered angle of the steered wheels 80 based on this converted value.

The second speed reducer 60 is connected to the steering motor 32 and the rack shaft 70 , and applies a force for steering the steered wheels 80 to the rack shaft 70 using the steering motor 32 as a drive source. The second speed reducer 60 reduces the rotational speed of the steering motor 32 and increases the rotational driving force to transmit it to the rack shaft 70 . It is converted into force and transmitted to the rack shaft 70 . For example, the second speed reducer 60 includes a first pulley 61 connected to the drive rotation shaft of the steering motor 32, a cylindrical second pulley 62 surrounding the outer circumference of the rack shaft 70, the first pulley 61 and the second An endless belt 63 stretched over pulleys 62 is provided. The second pulley 62 constitutes a ball nut. The spiral groove formed on the inner peripheral surface of the second pulley 62 and the spiral groove formed on the outer peripheral surface of the rack shaft 70 are screwed together via a plurality of balls (not shown). The second pulley 62 and the rack shaft 70 as described above constitute a ball screw mechanism. The rotational driving force of the steering motor 32 rotates the second pulley 62 via the first pulley 61 and the belt 63 , and the rotation of the second pulley 62 causes the rack shaft 70 to rotate relative to the second pulley 62 . It reciprocates in the axial direction. As a result, the rack shaft 70 steers the steerable wheels 80 . Note that the first pulley 61 and the second pulley 62 may be engaged so as to transmit the rotational driving force via gears formed on the respective outer peripheral surfaces without the belt 63 .

The steering angle control of the steered wheels 80 by the ECU 100 will be described with reference to FIG. FIG. 2 is a diagram showing an example of a functional configuration of the ECU 100 of FIG. 1. As shown in FIG. As described above, the ECU 100 uses the steering angle information from the steering angle sensor 50 and the vehicle speed information from the speed sensor 91 to calculate the target steering angle. At this time, the ECU 100 feeds back the turning angle of the steered wheels 80 detected via the turning angle sensor 71 (also referred to as "detected turning angle" or "actual turning angle") to the calculation of the target turning angle. By doing so, the calculated target steering angle is corrected, and the steering motor 32 is controlled using the corrected target steering angle.

As shown in FIG. 2 , the ECU 100 includes a steering angle calculation section 101 , a steering angle correction section 102 and a storage section 103 . The storage unit 103 enables storage and retrieval of various information. The storage unit 103 is realized by, for example, a storage device such as a ROM, a RAM, a semiconductor memory such as a flash memory, a hard disk drive, or an SSD (Solid State Disk). Although storage unit 103 is included in ECU 100 in the present embodiment, storage unit 103 may be arranged outside ECU 100 and connected to ECU 100 via a wire or radio.

The steering angle calculator 101 acquires the steering angle θ and the vehicle speed ν from the steering angle sensor 50 and the speed sensor 91 . The timing at which the vehicle speed ν is detected by the speed sensor 91 is at the same time, immediately after, or in the vicinity of the timing at which the steering angle θ is detected by the steering angle sensor 50 . Further, the steering angle calculation unit 101 determines the ratio of the input steering angle from the steering angle sensor 50 to the target steering angle (input steering angle/target steering angle) R based on the acquired vehicle speed ν. A steering angle δ is calculated. The steering angle calculator 101 outputs the target steering angle δ to the steering angle corrector 102 .

Here, the vehicle steering system 1 arbitrarily changes the steering angle ratio, which is the ratio of the steering angle/steering angle, to determine the target steering angle corresponding to the input steering angle from the steering angle sensor 50. can be done. In this specification and claims, the "steering angle ratio" is a ratio in which the denominator is the amount of change in steering angle and the numerator is the amount of change in steering angle. That is, the steering angle ratio is the ratio of steering angle/steering angle. Furthermore, hereinafter, the "input steering angle/target steering angle ratio R" is also referred to as the "target steering angle ratio R".

The target steering angle ratio R is determined in advance according to the vehicle speed. For example, the target steering angle ratio R is increased in a high-speed region where the vehicle speed is high, so that the response of the steering operation to the steering of the steering wheel 10 by the driver is dulled. In addition, the target steering angle ratio R is made small in a low vehicle speed range, and the response of the steering operation to the steering of the steering wheel 10 by the driver is sharpened. That is, the magnitude of the steering amount with respect to the steering amount of the steering wheel 10 is increased. The target steering angle ratio R may be set so as to change stepwise according to the vehicle speed, or may be set so as to change gradually in conjunction with the vehicle speed.

The target steering angle ratio R as described above is associated with each vehicle speed and stored in advance in the storage unit 103 . The steering angle calculation unit 101 acquires the target steering angle ratio R corresponding to the vehicle speed ν from the storage unit 103, and uses the acquired target steering angle ratio R to calculate the target steering angle δ. The turning angle calculator 101 calculates a target turning angle δ, a target turning angle ratio R used to calculate the target turning angle δ, and a vehicle speed ν used to calculate the target turning angle δ. , are associated with each other and output to the steering angle correction unit 102 .

The steering angle correction unit 102 acquires the target steering angle δ, the target steering angle ratio R, and the vehicle speed ν from the steering angle calculation unit 101 and acquires the actual steering angle δ ^* from the steering angle sensor 71 . do. Using the obtained target steering angle δ, target steering angle ratio R, and vehicle speed ν, the steering angle correction unit 102 calculates the corrected target steering angle δa with feedback based on the actual steering angle δ ^* . do. Such turning angle correction unit 102 includes acquisition unit 1021 , reference model calculation unit 1022 , virtual vehicle calculation unit 1023 , auxiliary steering angle calculation unit 1024 , and turning angle determination unit 1025 . The corrected target steering angle δa is an example of the output steering angle.

The acquisition unit 1021 acquires the target steering angle δ, the target steering angle ratio R, and the vehicle speed ν from the steering angle calculation unit 101 . Acquisition unit 1021 associates target turning angle δ and vehicle speed ν with each other and outputs them to turning angle determination unit 1025 . Acquisition unit 1021 also associates target steering angle δ, target steering angle ratio R, and vehicle speed ν with each other and outputs them to reference model calculation unit 1022 .

The reference model calculator 1022 calculates a yaw rate (also referred to as a “target yaw rate”) γm to be generated in the vehicle A using the acquired target steering angle δ and vehicle speed ν. The target yaw rate γm is an estimated yaw rate that can occur in the vehicle A when the vehicle A turns at the target steering angle δ and vehicle speed ν. Calculation of the target yaw rate γm from the target steering angle and vehicle speed can be performed using various known methods. For example, a vehicle model corresponding to vehicle A is stored in advance in storage unit 103 as a reference model. An example of the vehicle model is a transfer function showing the relationship between the target steering angle δ, vehicle speed ν, and target yaw rate γm of the vehicle A. Such a transfer function can be determined from the specifications of the vehicle A, experiments, and the like. The reference model calculation unit 1022 refers to the transfer function of the storage unit 103, and calculates the target yaw rate γm by inputting the target steering angle δ and the vehicle speed ν to the transfer function.

Furthermore, reference model calculation section 1022 calculates target yaw rate γm using target yaw rate γm, target steering angle δ corresponding to target yaw rate γm, and target steering angle ratio R corresponding to target steering angle δ. processing to reduce the high frequency gain of The gain corresponds to the ratio of the amplitude Ama of the signal of the target yaw rate γma after the gain change to the amplitude Am of the signal of the target yaw rate γm. Gain is expressed as 20log ₁₀ (Amplitude Ama/Amplitude Am). That is, the gain corresponds to the magnification of the amplitude Ama of the signal of the target yaw rate γm after the gain change to the amplitude Am of the signal of the target yaw rate γm.

FIG. 3A shows the relationship between the frequency of the signal of the target yaw rate γm on the horizontal axis and the gain on the vertical axis. Note that the horizontal axis is a logarithmic scale. FIG. 3B shows the relationship between the frequency of the signal of the target yaw rate γm on the horizontal axis and the phase delay on the vertical axis. Note that the horizontal axis is a logarithmic scale. The phase delay is the phase of the response delay in which the signal of the target yaw rate γma after gain change lags behind the signal of the target yaw rate γm. The phase lag is also the response delay of the signal of the target yaw rate γma with respect to the reception of the signal of the input steering angle θ.

As shown in FIG. 3A, a plurality of relationships between the frequency of the signal of the target yaw rate γm and the gain are set according to the target steering angle ratio R. Such a relationship is also called a gain characteristic. In the present embodiment, the three regions, ie, the high region where the value of the target steering angle ratio R is large, the low region where the value of the target steering angle ratio R is small, and the intermediate region between the high region and the low region, are described above. relationship is set. In any region, the gain is lowered in the high frequency region of the signal of the target yaw rate γm. Furthermore, the gain in the high frequency region is lowered as the target steering angle ratio R becomes smaller. Therefore, the smaller the target steering angle ratio R, the smaller the amount of steering operation with respect to the input to the steering wheel 10 in the high frequency region. Storage unit 103 stores in advance a plurality of gain characteristics as shown in FIG. 3A.

Also, FIG. 3B shows the relationship between the phase delay and the frequency corresponding to each of the high range, middle range and low range of the target yaw rate γm shown in FIG. 3A. In any region, the phase lag is small in the high frequency region of the signal of the target yaw rate γm. That is, in the high frequency region, the response of the steering execution timing to the input to the steering wheel 10 becomes sharp. Furthermore, the phase delay characteristics are the same in each of the high, middle and low regions. Therefore, regardless of the target steering angle ratio R, the response phase does not change. Then, for example, the amount of decrease in phase lag, which may increase as the amount of decrease in gain increases, is reduced. Note that the phase characteristics in each of the high, middle, and low regions may not be the same. Storage unit 103 stores a plurality of phase characteristics in advance.

In this way, the reference model calculation unit 1022 uses the target yaw rate γm, the corresponding target steering angle ratio R, and the relationships shown in FIGS. A signal of the target yaw rate γma is generated by reducing the high-frequency gain and reducing the phase delay of the response, and is output to the auxiliary steering angle calculation section 1024 . Hereinafter, the target yaw rate γma will be referred to as "second target yaw rate γma", and the target yaw rate γm will also be referred to as "first target yaw rate γm".

The virtual vehicle calculator 1023 acquires the actual turning angle δ ^* of the steered wheels 80 via the turning angle sensor 71 and the vehicle speed ν from the speed sensor 91 . The timing at which the actual steering angle δ ^* and the vehicle speed ν are detected by the steering angle sensor 71 and the speed sensor 91 is at the same time, immediately after, or in the vicinity of the timing at which the steering angle θ is detected by the steering angle sensor 50 . Using the acquired actual steering angle δ ^* and vehicle speed ν, the virtual vehicle calculation unit 1023 calculates the yaw rate (also called “estimated yaw rate”) γe and sideslip angle (also called “estimated sideslip angle”) occurring in vehicle A. ) is calculated, that is, estimated. The estimated yaw rate is a yaw rate that can occur in the vehicle A when the vehicle A turns at the actual turning angle δ ^* and the vehicle speed ν. The estimated sideslip angle is the angle between the traveling direction of the center of gravity of the vehicle A and the direction of the vehicle A (the direction of the center line of the vehicle). Calculation of the estimated yaw rate γe and the estimated sideslip angle β from the actual steering angle and vehicle speed can be performed using various known methods. For example, the estimated yaw rate γe can be calculated in the same manner as the calculation of the first target yaw rate γm. The virtual vehicle calculation unit 1023 refers to the transfer function indicating the vehicle model corresponding to the vehicle A stored in the storage unit 103, and inputs the actual turning angle δ ^* and the vehicle speed ν to the transfer function to obtain the estimated yaw rate Calculate γe. Virtual vehicle calculator 1023 outputs estimated yaw rate γe and estimated sideslip angle β to auxiliary steering angle calculator 1024 .

Using the second target yaw rate γma obtained from the reference model calculation unit 1022 and the estimated yaw rate γe and the estimated sideslip angle β obtained from the virtual vehicle calculation unit 1023, the auxiliary steering angle calculation unit 1024 calculates the target steering angle. A correction steering angle δc for correcting δ is calculated. Specifically, the auxiliary steering angle calculator 1024 multiplies the second target yaw rate γma by the feedback coefficient K3, multiplies the estimated yaw rate γe by the feedback coefficient K1, and multiplies the estimated sideslip angle β by the feedback coefficient K2. The auxiliary steering angle calculator 1024 calculates the corrected steering angle δc using the second target yaw rate γma multiplied by each feedback coefficient, the estimated yaw rate γe, and the estimated sideslip angle β. Then, the auxiliary steering angle calculation section 1024 outputs the corrected steering angle δc to the steering angle determination section 1025 . The correction steering angle δc can be calculated using various known methods. For example, the corrected steering angle δc can be calculated by the following formula.
δc=K3×γma+K1×γe+K2×β

In this embodiment, the feedback coefficients K1, K2 and K3 are calculated by the auxiliary steering angle calculator 1024. FIG. Specifically, the auxiliary steering angle calculator 1024 uses the output error e (e=γe−γma), which is the difference between the estimated yaw rate γe and the second target yaw rate γma, to calculate the evaluation function shown in Equation 1 below. Then, the feedback coefficients K1, K2 and K3 that minimize this evaluation function are calculated. In the following evaluation function, q is a weighting factor (constant). δ is the target steering angle corresponding to the input steering angle θ. The output error e corresponds to the error between the target steering angle δ and the actual steering angle δ ^* . Therefore, the error between the target steering angle δ and the actual steering angle δ ^* is fed back to the calculation of the corrected steering angle δc.

The turning angle determining unit 1025 adds the corrected steering angle δc acquired from the auxiliary steering angle calculating unit 1024 to the target turning angle δ acquired from the acquiring unit 1021, so that the target turning angle after correction is obtained. A certain corrected target steering angle δa is calculated. Corrected target steering angle δa=target steering angle δ+corrected steering angle δc. The ECU 100 controls the steering motor 32 so that the steered wheels 80 are steered at the corrected target steering angle δa determined by the steering angle determining section 1025 .

The ECU 100 of the vehicle steering system 1 according to Embodiment 1 as described above calculates the first target yaw rate γm based on the target turning angle δ corresponding to the input steering angle θ, and further calculates the first target yaw rate γm The second target yaw rate γma is calculated by lowering the high-frequency gain of . The ECU 100 also calculates an estimated yaw rate γe and an estimated sideslip angle β of the vehicle A based on the actual steering angle δ ^* . The ECU 100 calculates a corrected steering angle δc based on the second target yaw rate γma, the estimated yaw rate γe, and the estimated sideslip angle β, calculates a corrected target steering angle δa from the corrected steering angle δc and the target steering angle δ, and corrects The target steering angle δa is set as the target steering angle for controlling the steering motor 32 . The ECU 100 feeds back the error between the target steering angle δ and the actual steering angle δ ^* to the calculation of the corrected steering angle δc based on the second target yaw rate γma, the estimated yaw rate γe, and the estimated sideslip angle β. Steering control is performed with improved matching accuracy with respect to the target steering angle δ, which is input information.

As described above, the vehicle steering system 1 according to Embodiment 1 independently controls the output steering angle output to the steering mechanism 3 with respect to the input steering angle input from the steering wheel 10. . The ECU 100 of the vehicle steering system 1 determines a target steering angle ratio R that is a ratio of an input steering angle to a target steering angle, and calculates a target steering angle according to the input steering angle and the target steering angle ratio R. A steering angle calculator 101 and a steering angle corrector 102 for correcting a target steering angle and outputting it as an output steering angle are provided. The turning angle correction unit 102 corrects the target turning angle so that the smaller the target turning angle ratio R, the greater the suppression of the high frequency component of the target turning angle.

According to the above aspect, as the steering angle ratio becomes smaller, the reaction of the steering operation to the steering of the steering wheel 10 becomes more sensitive. Under such circumstances, the vehicle steering system 1 greatly suppresses the high-frequency component of the target steering angle, that is, the component with a high steering speed. Therefore, even if the driver steers the steering wheel 10 to a large turning angle such as when turning at an intersection, the vehicle A is prevented from overreacting and turning. Therefore, the driver does not need to change the grip of the steering wheel 10, such as gripping the steering wheel 10 again, in order to respond to the steering operation that reacts with hypersensitivity. Moreover, when the vehicle A runs at high speed, the stability of the vehicle A can be maintained even if the steering angle ratio is reduced. In addition, in the region where the steering angle is small, the component of high steering speed is greatly suppressed, so that the vehicle A is prevented from swaying when traveling straight ahead. Furthermore, even if the steering angle ratio is reduced in the entire vehicle speed range, the vehicle A runs stably, so it is possible to reduce the operation amount of the steering wheel 10 in the entire vehicle speed range. Thus, the vehicle steering system 1 can improve the ease of driving for the driver. Such a vehicle steering system 1 can reduce the influence of changes in the steering angle ratio on running.

Further, in the vehicle steering system 1 according to Embodiment 1, the turning angle correction unit 102 equalizes the phase delay of the output turning angle with respect to the target turning angle between different target turning angle ratios R. FIG. According to the above aspect, the response delay in the high frequency region of the steering operation to the steering input to the steering wheel 10 is the same regardless of the target steering angle ratio R. Therefore, even if the steering angle ratio changes, the driver can steer without discomfort.

Further, in the vehicle steering system 1 according to Embodiment 1, the target steering angle ratio R decreases as the speed of the vehicle A decreases. According to the above aspect, as the speed of the vehicle A decreases, the driver can turn the vehicle A at a large turning angle with a small amount of steering. As the speed of the vehicle A decreases, the vehicle A requires a larger steering angle for turning, so it becomes easier for the driver to drive the vehicle A.

Further, in the vehicle steering system 1 according to Embodiment 1, the turning angle correction unit 102 acquires the actual turning angle, which is the actual turning angle in the turning mechanism 3, and calculates the target turning angle and the actual turning angle. A correction angle to be given to the target steering angle is calculated using the steering angle, and the error between the target steering angle and the actual steering angle is fed back to the calculation of the correction angle. According to the above configuration, the correction angle is calculated using the target turning angle and the actual turning angle, so the state of the vehicle A can be reflected. That is, the output steering angle can reflect the state of the vehicle A. Furthermore, since the correction angle receives feedback of the error, the accuracy of the output steering angle with respect to the target steering angle is improved. Therefore, accurate steering operation becomes possible.

Further, in the vehicle steering system 1 according to Embodiment 1, the reference model calculation unit 1022 uses the first target yaw rate γm, the target steering angle δ, and the target steering angle ratio R to calculate the high-frequency model of the first target yaw rate γm. Although the processing for lowering the gain has been performed, it is not limited to this. For example, the reference model calculation unit 1022 may use the target steering angle δ and the target steering angle ratio R to perform processing to reduce the high-frequency gain of the target steering angle δ. Again, the relationships shown in FIGS. 3A and 3B are used. Specifically, the horizontal axis in the diagrams corresponding to FIGS. 3A and 3B indicates the frequency of the signal of the target steering angle δ. The vertical axis in the drawing corresponding to FIG. 3A indicates the gain of the target steering angle δ. The vertical axis in the diagram corresponding to FIG. 3B indicates the phase delay of the signal of the target steering angle δ after the gain change. Then, reference model calculation section 1022 calculates the second target yaw rate γma using the target steering angle after the high-frequency gain is lowered.

[Embodiment 2]
A vehicle steering system according to Embodiment 2 will be described. In the vehicle steering system according to Embodiment 2, the ECU 200 calculates the target yaw rate from the target turning angle using a filter model constructed by machine learning. In the following, the points that are different from the first embodiment will be mainly described, and the description of the points that are the same as the first embodiment will be omitted.

FIG. 4 shows an example of the functional configuration of the ECU 200 in the vehicle steering system according to the second embodiment. As shown in FIG. 4 , the ECU 200 includes a steering angle calculation section 101 , a steering angle correction section 202 and a storage section 103 . Turning angle correction section 202 includes acquisition section 2021 , filter section 2027 , and turning angle determination section 2025 . Acquisition unit 2021 acquires target turning angle δ and vehicle speed ν corresponding thereto from turning angle calculation unit 101 .

Acquisition section 2021 outputs target steering angle δ to filter section 2027 and outputs vehicle speed ν to steering angle determination section 2025 . The filter unit 2027 calculates the second target yaw rate γma using the target turning angle δ and outputs it to the turning angle determination unit 2025 . Furthermore, the filter unit 2027 causes the storage unit 103 to store the target steering angle δ. As a result, in the storage unit 103, a plurality of target steering angles acquired so far are associated with the order in which they were acquired, and are accumulated, for example, as time-series data. Note that the storage of the target steering angle δ in the storage unit 103 may be performed by the steering angle calculation unit 101 or the acquisition unit 2021 . A steering angle determination unit 2025 calculates a corrected target steering angle δa by correcting the target steering angle δ using the vehicle speed ν and the second target yaw rate γma. For example, the steering angle determining unit 2025 calculates the corrected target steering angle δa from the vehicle speed ν and the second target yaw rate γma by performing inverse calculation of the transfer function of the reference model described in the first embodiment.

Details of the filter unit 2027 will be described. Filter unit 2027 outputs output data for input data using a filter model. The filter model is a nonlinear filter model, and is a nonlinear time-series model that uses the target turning angle δ and past values of the target turning angle δ as input data and the second target yaw rate γma as output data. The past value of the target steering angle δ is a target steering angle acquired before the target steering angle δ and accumulated in the storage unit 103 . A filter model is stored in the storage unit 103 . The filter unit 2027 performs high-frequency gain reduction processing and response phase delay suppression processing according to the target steering angle ratio R, for example, as in Embodiment 1, on the input data as described above. Output the second target yaw rate γma. Note that the time series model is a model that gives rules between variables between two points in time series data. Time-series data is data that is observed or acquired in order over time. For example, the steering angle detected by the steering angle sensor 50 is time-series data, and the target turning angle calculated from each steering angle is time-series data.

In this embodiment, the filter model is a nonlinear autoregressive time series model. The filter model improves the output accuracy of the output data by learning using the learning data. A learning model applied to the filter model is a machine learning model using a neural network such as Deep Learning, but may be another learning model. For example, the learning model may be a machine learning model using Random Forest, Genetic Programming, or the like.

The machine learning of the filter model may be performed by the ECU 200, or may be performed by an external device such as a server device (not shown) arranged at a position remote from the ECU 200. FIG. A filter model learned by an external device such as a server device may be sent to the storage unit 103 by wireless communication via a recording medium such as a nonvolatile memory or a communication network such as the Internet.

Supervised learning (also called “supervised learning”) is used to train the filter model. In supervised learning, the weightings in the neural network are adjusted so that the input data, ie the output of the filter model for the input signal, matches the teacher signal. In this embodiment, in the neural network, the current value y(t) is obtained as a function of the past data of each of the teacher signal y(t) and the input signal x(t) as shown in Equation 2 below: is expected. Note that f is a nonlinear function. The learning of the filter model is performed using a teacher signal created in advance, and after the learning is completed, the output y(t) from the filter model is connected to the input y(t-1), and a closed loop neural network and be done.

The details of the training method of the filter model will be explained. FIG. 5 shows an example of a configuration for learning the filter model 2027a according to the second embodiment. As described above, the configuration of FIG. 5 may be included in ECU 200 or may be included in an external device such as a server device. When included in ECU 200, filter unit 2027A in FIG. 5 is filter unit 2027 in FIG. The learning of the filter model 2027a is performed by the filter section 2027A and the teacher signal generation section 2027B. The filter section 2027A includes a filter model 2027a and a vehicle model 2027b, and the teacher signal generation section 2027B includes a vehicle model 2027b and a zero phase filter (Zero Phase Filter) 2027c.

The filter model 2027a performs high-frequency gain reduction processing and response phase delay suppression processing according to the target steering angle ratio R on the input data. Therefore, as shown in FIG. 5, a zero-phase filter 2027c is used to generate the teacher signal. The zero-phase filter 2027c is an off-line filter that can zero the phase delay of the output by using not only the past value of the input data but also the future value. Using the phase characteristic that the phase delay of the zero-phase filter 2027c can be made zero, it is possible to create a teacher signal whose gain and phase can be changed independently. Note that "online" indicates that the vehicle is running, and "offline" indicates that the vehicle is not running. The off-line filter uses pre-acquired data and data acquired when the vehicle is not running, such as artificially generated data.

Offline filters can use future values, while online filters can only use past values. When the filter model 2027a is actually used, only past values can be used because it is online. Therefore, by using the nonlinearity of the neural network, the filter model 2027a can pseudo-realize characteristics close to those of the zero-phase filter 2027c while being an online filter. By using the filter model 2027a learned offline using the teacher signal, even the online filter can simulate characteristics close to those of the teacher signal.

When an input signal such as a target steering angle is input, the teacher signal generation section 2027B inputs the input signal to the vehicle model 2027b and outputs it as a yaw rate signal from the vehicle model 2027b. The vehicle model 2027b is a model representing the characteristics of the target vehicle, and is, for example, a transfer function representing the vehicle model in the first embodiment. Further, the teacher signal generator 2027B inputs the yaw rate signal to the zero phase filter 2027c, and outputs the yaw rate signal as the teacher signal from the zero phase filter 2027c. Such teacher signals are pre-computed and used for training the filter model 2027a. The teacher signal is not limited to a teacher signal generated from data regarding one vehicle, and may be, for example, a general-purpose teacher signal generated from data regarding a plurality of vehicles having similar characteristics.

In learning the filter model 2027a, the filter unit 2027A inputs an input signal such as a target steering angle to the vehicle model 2027b, and inputs a yaw rate signal output from the vehicle model 2027b to the filter model 2027a. The filter unit 2027A also inputs the input signal to the filter model 2027a without going through the vehicle model 2027b. The input signal includes a target steering angle and a past value of the target steering angle. An example of the input signal is time-series data of the target steering angle. Therefore, the filter model 2027a receives the target steering angle, the past values (x1 to xn) of the target steering angle, and the yaw rate generated from these target steering angles. The signals of the input target steering angle and yaw rate are output from the filter model 2027a as a yaw rate signal. The filter unit 2027A adjusts weighting between nodes in the neural network forming the filter model 2027a so that the yaw rate signal output from the filter model 2027a matches the teacher signal.

A neural network is composed of a plurality of node layers including an input layer and an output layer. The node layer contains one or more nodes. For example, when a neural network is composed of three node layers, an input layer, an intermediate layer, and an output layer, the neural network performs output processing from the input layer to the intermediate layer for information input to the nodes of the input layer, The processing in the intermediate layer, the output processing from the intermediate layer to the output layer, and the processing in the output layer are sequentially performed to output an output result that matches the input information. Each node in one layer is connected to each node in the next layer, and the connections between nodes are weighted. The information of the nodes of one layer is weighted by the connection between nodes and output to the nodes of the next layer.

Also, the filter model 2027a is constructed so as to correspond to a plurality of gain characteristics. Therefore, the zero-phase filter 2027c is created so as to correspond to a plurality of gain characteristics without changing the phase characteristics. A natural frequency corresponding to the vehicle is set in the vehicle model 2027b. A natural frequency corresponding to each gain characteristic is set in the zero-phase filter 2027c. Such a zero-phase filter 2027c exhibits frequency responses of a plurality of vehicle specifications having the same phase characteristics and different gain characteristics. The frequency response is the frequency response of yaw rate to steering angle. As described with reference to FIGS. 3A and 3B in Embodiment 1, when three gain characteristics of the high region, middle region and low region are considered, the zero phase filter 2027c has natural vibration corresponding to the three gain characteristics. number can be set. The filter model 2027a constructed as described above can calculate an appropriate second target yaw rate γma corresponding to the target steering angle δ even if the relationship between gain and phase delay is nonlinear.

Filter model 2027 a learned as described above is stored in storage unit 103 . Then, the filter unit 2027 of the turning angle correction unit 202 of the ECU 200 converts the target turning angle δ acquired while the vehicle A is running and the past value of the target turning angle δ acquired from the storage unit 103 into By inputting to the vehicle model stored in the storage unit 103, the target steering angle δ and the yaw rate corresponding to its past value are obtained. The vehicle model is similar to the vehicle model 2027b. Furthermore, the filter unit 2027 inputs the target turning angle δ, its past value, and the yaw rate into the filter model 2027a of the storage unit 103, thereby calculating the second target yaw rate γma and sending it to the turning angle determination unit 2025. Output. Data consisting of the target steering angle δ and its past value correlates with the target steering angle ratio R. Such a filter unit 2027 outputs the second target yaw rate γma that reflects the decrease in the high-frequency gain according to the target steering angle ratio R and the suppression of the response phase delay.

Other configurations and operations of the vehicle steering system according to Embodiment 2 are the same as those in Embodiment 1, and therefore descriptions thereof are omitted. Further, according to the vehicle steering system according to the second embodiment, the same effects as those of the first embodiment can be obtained.

Further, in the ECU 200 of the vehicle steering system according to the second embodiment, the steering angle correction unit 202 uses a nonlinear filter model to correct the target steering angle so as to suppress high-frequency components. According to the above configuration, it is possible to appropriately suppress high frequency components in accordance with various relationships between gain and phase delay. For example, if the gain and the phase lag are in a linear relationship, the smaller the phase lag, the larger the gain at high frequencies.

In ECU 200 of the vehicle steering system according to Embodiment 2, the nonlinear filter model is a time-series model learned using a neural network. By using neural networks, it is possible to construct appropriate nonlinear filter models that correspond to various gain and phase lag relationships.

[others]
Although the vehicle steering system according to one or more aspects of the present invention has been described above based on the embodiments, the present invention is not limited to the embodiments. As long as it does not deviate from the spirit of the present invention, various modifications that a person skilled in the art can think of, and a form constructed by combining the components of different embodiments are also one or more aspects of the present invention. may be included within the range of

For example, the steering angle correction unit 202 of the vehicle steering system according to Embodiment 2 may include a filter unit having a low-pass filter instead of the filter unit 2027 having the filter model 2027a. The low-pass filter may be a filter that removes high-frequency components of the target steering angle or high-frequency components of the target yaw rate calculated from the target steering angle. Furthermore, the low-pass filter may change the high-frequency component to be removed according to the target steering angle ratio R, for example, as in the first embodiment. This also makes it possible to reduce the influence of changes in the steering angle ratio on running.

Further, although the vehicle steering system according to the embodiment has been described as constituting a steer-by-wire system, it is not limited to this. The vehicle steering system may be any steering system having a variable steering angle ratio. For example, the vehicle steering system may have a configuration in which the steering mechanism and the steering mechanism are mechanically connected, and a variable mechanism for changing the ratio is provided between the steering mechanism and the steering mechanism. . Alternatively, the vehicle steering system may constitute a steer-by-wire system capable of left and right independent steering. In the steer-by-wire system, the steering mechanisms for the left and right steerable wheels are not mechanically coupled to each other, nor are they mechanically coupled to the steering mechanism. The left and right steering mechanisms are operated by actuators provided in each steering mechanism.

In addition, all numbers such as ordinal numbers and numbers used above are examples for specifically describing the present invention, and the present invention is not limited to the illustrated numbers. Moreover, the connection relationship between the components is an example for specifically describing the present invention, and the connection relationship for realizing the function of the present invention is not limited to this.

Also, the division of functional blocks in the block diagram is an example, and a plurality of functional blocks can be realized as one functional block, one functional block can be divided into a plurality of functional blocks, and some functions can be moved to other functional blocks. may Moreover, single hardware or software may process the functions of a plurality of functional blocks having similar functions in parallel or in a time-sharing manner.

INDUSTRIAL APPLICABILITY The vehicle steering system according to the present invention is useful as a vehicle steering system having a variable steering angle ratio.

1 Vehicle Steering Device 2 Steering Mechanism 3 Steering Mechanism 10 Steering Wheel 32 Steering Motor 50 Steering Angle Sensor 71 Steering Angle Sensor 80 Steering Wheel 91 Speed Sensor 100, 200 ECU 101 Turning steering angle calculation unit 102, 202 steering angle correction unit A vehicle

Claims (6)

A vehicle steering apparatus that independently controls an output steering angle output to a steering mechanism with respect to an input steering angle input from a steering wheel,
a steering angle calculation unit that determines a ratio of the input steering angle to a target steering angle and calculates the target steering angle according to the input steering angle and the ratio;
a turning angle correction unit that corrects the target turning angle and outputs it as the output turning angle;
The steering angle correction unit corrects the target steering angle such that the smaller the ratio, the greater the suppression of high frequency components of the target steering angle. 2. The vehicle steering system according to claim 1, wherein the turning angle correcting section equalizes the phase delay of the output turning angle with respect to the target turning angle between the different ratios. 3. The vehicle steering system according to claim 1, wherein the ratio decreases as the speed of the vehicle decreases. The steering angle correction unit is
Acquiring an actual steering angle, which is an actual steering angle in the steering mechanism,
calculating a correction angle given to the target steering angle using the target steering angle and the actual steering angle;
4. The vehicle steering system according to any one of claims 1 to 3, wherein an error between said target turning angle and said actual turning angle is fed back to calculation of said correction angle. The vehicle steering system according to any one of claims 1 to 3 , wherein the turning angle correction unit uses a nonlinear filter model to correct the target turning angle so as to suppress high-frequency components. The vehicle steering system according to claim 5, wherein the nonlinear filter model is a time-series model learned using a neural network.

Images