JP6273706B2 - Motor control device - Google Patents

Description

  The present invention relates to a technique for controlling a motor that generates assist torque and automatic steering torque.

  Based on the image information from the camera that captures the front of the vehicle, it detects the positional relationship between the running lane and the hourly vehicle, and assists steering operation by the driver with lane keeping control that realizes traveling along the lane A device that realizes power steering control for generating assist torque with a single actuator (motor) is known (see, for example, Patent Document 1).

  In this apparatus, basically, the motor is driven based on the result of adding the necessary torque for lane keeping control to the necessary torque for power steering. However, when lane keep control is being performed, the influence of power steering control is suppressed by multiplying the required torque for power steering by a coefficient greater than 0 and less than 1 so that it does not easily deviate from the lane. doing.

JP-A-9-221053

  By the way, when an intervention operation by the driver during lane keep control, so-called driver override, is performed, the deviation of the actual position / actual angle with respect to the target position / angle set in the lane keep control increases. Torque is generated to try to cancel out the enlarged deviation.

  In addition, the lane keeping control is more responsive than the assist control (for example, while the assist control is on the order of 0.1 Hz, the lane keeping control is required to ensure robustness against road disturbances and sudden steering. The control requires 1 to 100 Hz order).

  However, if the lane keep control is highly responsive, a large torque that instantly cancels the deviation caused by the driver override is generated and hinders the operation of the driver, which causes the driver to feel uncomfortable. there were.

  In addition, when the driver is steering and system intervention (for example, target tracking control for avoiding danger) is performed, if the target tracking control has a high response, the target value of the target tracking control is set. There is a problem that the driver feels uncomfortable due to the sudden automatic steering.

  In order to solve such a problem, in the control of the motor that generates the assist torque and the automatic steering torque, switching of the control is realized without causing the driver to feel strange when the driver or the system intervenes with respect to the current control. For the purpose.

  The motor control device of the present invention includes assist control means, follow-up control means, motor drive means, intervention detection means, and restriction means. The assist control means generates an assist command for generating an assist torque that reduces the steering load in accordance with the detected value of the steering torque. The follow-up control means acquires a target value of a physical quantity related to steering, and generates a follow-up command for generating an automatic steering torque that causes the detected value of the physical quantity to follow the target value. The motor driving means drives the motor that generates the assist torque and the automatic steering torque according to the added value of the assist command and the follow-up command.

  The intervention detection means detects intervention by the driver or the system. The limiting means includes at least one of the assist control means and the follow-up control means as a target means, and is provided in the target means. The ratio of the assist torque to the automatic steering torque is provided in the target means according to the degree of intervention detected by the intervention detection means. Limit the internal values used by the target means.

  Normally, when an intervention operation (driver override) by a driver is detected during the execution of the tracking control by the tracking control means, the tracking control means performs an intervention operation in order to cancel the steering torque against the automatic steering torque generated by the intervention operation. Generates a large follow-up command that resists For this reason, the higher the response of the target tracking control, the more difficult the intervention operation is.

  On the other hand, in the present invention, when the intervention operation by the driver is detected, the ratio of the assist torque to the automatic steering torque is increased, so that generation of a large follow-up command against the intervention operation can be suppressed. it can. As a result, when the driver performs an intervention operation, the follow-up control can be shifted to the assist control without causing the driver to feel uncomfortable.

  In addition, in the target tracking control, an integration function is usually included to make the detected value coincide with the target value, and when an intervention operation is performed, the integration of the internal value proceeds rapidly to counter this. . Generally, the upper limit of the follow-up command is used to protect the motor and the motor drive circuit. However, the integration of the internal value proceeds without being restricted, and so-called widening occurs. In other words, even if an intervention operation is performed in a direction that decreases the integral value, the command value for the upper limit value (and thus excessive automatic steering torque) continues to be output until the integral value falls below the upper limit value for the follow-up command. This gives the driver a feeling of strangeness.

  On the other hand, in the present invention, since the integration of the internal value can be limited, the occurrence of windup during the intervention operation can be suppressed, and the steering intended by the driver can be realized quickly. . However, the integration function here is not limited to a function realized by an integration circuit, but also includes a function realized by a low-pass filter and a transfer function including a 1 / s term.

  In addition, the code | symbol in the parenthesis described in the claim shows the correspondence with the specific means as described in embodiment mentioned later as one aspect, Comprising: The technical scope of this invention is limited is not.

  In addition to the motor control device described above, the present invention can be implemented in various forms such as a system including the motor control device as a component, a program for causing a computer to function as each means constituting the motor control device, and a motor control method. Can be realized.

1 is an overall configuration diagram of an electric steering system according to a first embodiment. It is a block diagram which shows schematic structure of EPS-ECU. It is a flowchart which shows the processing content of an intervention detection part. It is a block diagram which shows the structure of a target tracking control calculating part. It is a flowchart which shows the processing content of a limiting calculator. It is a wave form diagram which shows the operation example of a limiting calculator. It is a wave form diagram which shows the operation example of the limiting calculator in the modification of 1st Embodiment. It is a block diagram which shows the structure of the target tracking control calculating part in 2nd Embodiment. It is a flowchart which shows the processing content of the limiting calculator in 2nd Embodiment. It is a block diagram which shows the structure of the target tracking control calculating part in 3rd Embodiment. It is a flowchart which shows the processing content of the limiting calculator in 3rd Embodiment. It is a block diagram which shows the structure of the target tracking control calculating part in 4th Embodiment. It is a flowchart which shows the processing content of the limiting calculator in 4th Embodiment. It is a block diagram which shows the structure of the target tracking control calculating part in 5th Embodiment. It is a flowchart which shows the processing content of the limiting calculator in 5th Embodiment. It is a block diagram which shows the structure of the target tracking control calculating part in 6th Embodiment. It is a flowchart which shows the processing content of the limiting calculator (prefix limiting calculator) in 6th Embodiment. It is a flowchart which shows the processing content of the limiting calculator (postfix limiting calculator) in 6th Embodiment. It is a block diagram which shows the structure of the target tracking control calculating part in 7th Embodiment. It is a flowchart which shows the processing content of the limiting calculator in 7th Embodiment. It is a block diagram which shows schematic structure of EPS-ECU in 8th Embodiment. It is a graph which shows the conversion characteristic from the steering torque in the intervention detection part in 8th Embodiment to the intervention coefficient. It is a block diagram which shows the structure of the assist control calculating part in 8th Embodiment. It is a flowchart which shows the processing content of the limiting calculator in 8th Embodiment. It is a block diagram which shows the structure of the assist control calculating part in the modification of 8th Embodiment. It is a block diagram which shows schematic structure of EPS-ECU in 9th Embodiment. It is a block diagram which shows the structure of the assist control calculating part in 9th Embodiment. It is a flowchart which shows the processing content of the limiting calculator in 9th Embodiment. It is a block diagram which shows the structure of the intervention detection part periphery in 10th Embodiment. It is a flowchart which shows the processing content of the intervention detection part in 10th Embodiment. It is a block diagram which shows the structure of the intervention detection part periphery in 11th Embodiment. It is a flowchart which shows the processing content of the intervention detection part in 11th Embodiment. It is a flowchart which shows the processing content of the intervention detection part in 12th Embodiment. It is a flowchart which shows the processing content of the intervention detection part in 13th Embodiment.

Embodiments of the present invention will be described below with reference to the drawings.
[First Embodiment]
<Overall configuration>
As shown in FIG. 1, the electric steering system 1 according to the present embodiment includes assist control for assisting a driver (operation of a steering wheel) 2 with a motor 6 and automatic steering along a target course set in a travel lane. The target follow-up control (in this case, lane keep control) is executed by the motor 6.

  The handle 2 is fixed to one end of a steering shaft 3, and a torque sensor 4 is connected to the other end of the steering shaft 3, and an intermediate shaft 5 is connected to the other end of the torque sensor 4. In the following description, the entire shaft body from the steering shaft 3 through the torque sensor 4 to the intermediate shaft 5 is also collectively referred to as a steering shaft.

  The torque sensor 4 is a sensor for detecting the steering torque Ts. Specifically, a torsion bar that connects the steering shaft 3 and the intermediate shaft 5 is provided, and a torque applied to the torsion bar is detected based on a twist angle of the torsion bar.

  The motor 6 is for generating an assist torque based on the assist control and an automatic steering torque based on the target follow-up control, and its rotation is transmitted to the intermediate shaft 5 via the speed reduction mechanism 6a. That is, the speed reduction mechanism 6a is constituted by a worm gear provided at the tip of the rotating shaft of the motor 6 and a worm wheel provided coaxially with the intermediate shaft 5 in mesh with the worm gear. The rotation of the motor 6 is transmitted to the intermediate shaft 5. Conversely, when the intermediate shaft 5 is rotated by the operation of the handle 2 or the reaction force from the road surface (road surface reaction force), the rotation is transmitted to the motor 6 via the speed reduction mechanism 6a, and the motor 6 also rotates. .

  The motor 6 is formed of, for example, a brushless motor, and includes a rotation sensor such as a resolver. The rotation sensor outputs at least the motor rotation angle θ and the motor rotation angular velocity ω. However, instead of the motor rotation angle θ and the motor rotation angular velocity ω, a steering angle and a steering angular velocity obtained by multiplying the motor rotation angle θ and the motor rotation angular velocity ω by the gear ratio of the speed reduction mechanism 6a may be used.

  The other end of the intermediate shaft 5 opposite to the end to which the torque sensor 4 is connected is connected to the steering gear box 7. The steering gear box 7 is configured by a gear mechanism including a rack and a pinion gear, and the rack teeth mesh with a pinion gear provided at the other end of the intermediate shaft 5. Therefore, when the driver turns the handle 2, the intermediate shaft 5 rotates (that is, the pinion gear rotates), thereby moving the rack to the left and right. Tie rods 8 are attached to both ends of the rack, and the tie rods 8 reciprocate left and right together with the rack. Accordingly, the tie rod 8 pulls or pushes the knuckle arm 9 ahead, thereby changing the direction of each tire 10 that is a steered wheel.

A vehicle speed sensor 11 for detecting the vehicle speed V is provided at a predetermined part of the vehicle.
Hereinafter, the entire mechanism from the steering wheel 2 to each tire 10 to which the steering force of the steering wheel 2 is transmitted is also collectively referred to as a steering system mechanism 100.

  In the steering system mechanism 100 having such a configuration, when the steering wheel 2 is rotated by the driver's steering, the rotation is transmitted to the steering gear box 7 via the steering shaft 3, the torque sensor 4, and the intermediate shaft 5. Then, in the steering gear box 7, the rotation of the intermediate shaft 5 is converted into the left-right movement of the tie rod 8, and the left and right tires 10 are steered by the movement of the tie rod 8.

The LKP (lane keep) -ECU 16 is operated by electric power from a vehicle battery (not shown), detects a vehicle lane and a position of the vehicle in the vehicle lane from an image in front of the vehicle captured by a vehicle camera (not shown), and detects the detection. Set the target course based on the results. Furthermore, based on the detected value or the like of the vehicle speed and the steering angle, set the target angle theta ^* is a target value of the motor rotational angle for traveling along the target course (or the steering angle), the target angle theta ^* Output to EPS-ECU 15. Since the process for setting the target angle θ ^* is well known in the lane keep control, the description thereof is omitted here.

The EPS (Electric Power Steering) -ECU 15 is operated by electric power from a vehicle battery (not shown), the target angle θ ^* obtained by the LKP-ECU 16, the steering torque Ts detected by the torque sensor 4, and the motor from the motor 6. Based on the rotation angle θ, the motor rotation angular velocity ω, and the vehicle speed V detected by the vehicle speed sensor 11, an assist command AC that is a current command value for generating an assist torque and a current command for generating an automatic steering torque. A final command TL is calculated by adding the follow-up command TC as a value. Then, by applying a drive voltage Vd corresponding to the final command TL to the motor 6, an assist torque and an automatic steering torque are generated.

That is, the EPS-ECU 15 controls the steering characteristics by controlling the motor 6 with the drive voltage Vd, and thus controls the steering system mechanism 100 driven by the motor 6.
<EPS-ECU>
As shown in FIG. 2, the EPS-ECU 15 detects an assist control calculation unit 20 that generates an assist command AC, a target tracking control calculation unit 30 that generates a tracking command TC, and an intervention operation to the target tracking control by a driver. To the motor 6 based on the drive command DC, an adder 50 that generates a drive command DC that becomes a current command value for driving the motor by adding the assist command AC and the follow-up command TC. A motor drive circuit 60 that energizes and drives the motor 6 by applying a drive voltage Vd (not shown, but applied for three phases if a three-phase motor) is provided.

  The assist control calculation unit 20, the target follow-up control calculation unit 30, the intervention detection unit 40, and the adder 50 are actually realized by a CPU (not shown) included in the EPS-ECU 15 executing a predetermined control program. . Here, in order to ensure the responsiveness required for the target tracking control (lane keep control), there is no problem in executing the LKP in the above control program at any cycle (for example, from several hundred us to several hundred ms). The drive command DC is updated at this cycle. However, the realization of these units by software is merely an example, and at least a part of them may be realized by hardware such as a logic circuit.

<< Motor drive circuit >>
Based on the drive command DC, the motor drive circuit 60 applies a drive voltage Vd to the motor 6 so that torque (assist torque and automatic steering torque) corresponding to the drive command DC is applied to the steering shaft. Specifically, the drive command DC is used as a target current, and the drive voltage Vd is feedback-controlled so that the energization current Im flowing through the motor 6 matches the target current, thereby generating a desired torque for the steering shaft. In addition, since such a motor drive circuit 60 is a well-known technique (for example, refer Unexamined-Japanese-Patent No. 2013-52793), the description about the detail is abbreviate | omitted.

<< Assist control calculation section >>
Based on the steering torque Ts, the motor rotation angular velocity ω, and the vehicle speed V, the assist control calculation unit 20 is configured so that a feeling of transmission corresponding to the road surface reaction force (road surface load) and a feel corresponding to the steering state are realized. An assist command AC representing a current command value for generating an assist torque for assisting the operation is generated. Specifically, for example, a basic assist amount for obtaining a sense of transmission according to the road surface reaction force is calculated based on the steering torque Ts and the vehicle speed V, and according to the steering state according to the steering torque Ts and the motor rotational angular velocity ω. An assist command AC is generated by calculating an assist compensation amount and adding the assist compensation amount multiplied by a gain corresponding to the vehicle speed V to the basic assist amount. However, the calculation method of the assist command AC is not limited to this, and any known method can be used.

<< Intervention detection part >>
As shown in FIG. 3, the intervention detection unit 40 first reads the steering torque Ts (S110). At this time, a low-pass filter (LPF) may be applied to the steering torque Ts to remove noise other than the intervention operation by the driver, such as a road surface disturbance superimposed on the steering torque Ts.

  Next, according to a conversion table prepared in advance, an intervention coefficient α corresponding to the read absolute value | Ts | of the steering torque Ts is calculated (S120). The intervention detection unit 40 supplies the calculated intervention coefficient α to the target follow-up control calculation unit 30.

  Note that the conversion table outputs α = 1 when | Ts | ≦ A, outputs α = 0 when | Ts | ≧ B, and increases | Ts | when A <| Ts | <B, α = It is set to output a monotonically decreasing value in the range of 1 to α = 0. That is, the intervention detection unit 40 generates an intervention coefficient α that decreases as the absolute value | Ts | of the steering torque increases, that is, as the steering intervention level of the driver increases. The range of | Ts | ≦ A is a dead zone, so that when the driver touches the handle unintentionally, this is not erroneously detected as an intervention operation. B is set to a magnitude that can be reliably determined to be the steering torque generated by an intentional operation, for example. Here, the absolute value | Ts | of the steering torque is used as the parameter of the conversion table, but the steering torque Ts itself may be used. In this case, the conversion table shows symmetrical characteristics across the axis of Ts = 0. In addition, when A <| Ts | <B, if α is decreased, the decreasing tendency is not limited to a linear one. For example, a form such as a quadratic function is used, or any known method is used. May be used.

<< Target tracking control calculation section >>
As shown in FIG. 4, the target follow-up control calculation unit 30 is necessary to make the actual angle θ follow the target angle θ ^* based on the target angle θ ^* and the motor rotation angle (hereinafter also referred to as “real angle”) θ. A follow-up command TC representing a current command value for generating a large automatic steering torque is generated. Specifically, the target follow-up control calculation unit 30 determines a control characteristic by adding a PID gain to the subtractor 31 that obtains the deviation of the actual angle θ from the target angle θ ^* and the deviation obtained by the subtractor 31. By limiting the internal value used for the calculation in the target follow-up control calculation unit 30 in accordance with the determiner 32, the integrator 33 that integrates the integration target value TM that is the output of the characteristic determiner 32, and the intervention coefficient α. And a limit calculator 34 for limiting the follow-up command TC. The characteristic determiner 32 and the integrator 33 are well-known ones that realize a control structure based on a mathematical expression obtained by performing a bilinear transformation for discretization of a mathematical expression representing general PID control. is there.

  The limit calculator 34 sets the output of the adder that adds the integration target value TM and the previous value TC [n−1] of the follow-up command in the integrator 33 as the limit target value u, and the output y after the limit is the follow-up command TC. [N] is connected.

  Then, as shown in FIG. 5, the limit calculator 34 first reads the limit target value u (output of the subtractor 31), reads the intervention coefficient α (S210), and intervenes in the preset upper limit CL of the follow-up command. The guard value LM is calculated by multiplying the coefficient α (S220). The upper limit value CL is set to the rated current of the motor 6, for example.

  Next, it is determined whether or not the absolute value | u | of the restriction target value is greater than or equal to the guard value LM (S230). That is, the range of −LM to LM is set as the allowable range of the follow-up command TC (guard value ± LM is a boundary value of the allowable range), and it is determined whether or not the restriction target value u exceeds the allowable range.

When the restriction target value u is within the allowable range (| u | <LM) (S230-NO), the restriction target value u is set as the output y without being restricted (S240).
When the restriction target value u is outside the allowable range (| u | ≧ LM) (S230-YES), it is determined whether the restriction target value u is a non-negative value (S250).

When the restriction target value u is a non-negative value (S250-YES), the guard value (the upper limit value of the allowable range) LM is set as the output y (S260).
When the restriction target value u is a negative value (S250-NO), the negative guard value (lower limit value of the allowable range) -LM is set as the output y (S270).

<Operation>
In the electric steering system 1 configured as described above, as shown in FIG. 6, when the target follow-up control (lane keep control) is being performed, the follow-up command TC is a non-zero value (usually, | TC | <LM On the other hand, the steering torque Ts and the assist command AC are maintained at Ts = 0 and AC = 0 unless the driver intervenes. During this time, since | Ts | <A, the intervention coefficient α is 1 (the allowable range of the follow-up command TC is maximum).

  When the driver intervenes in the target follow-up control by operating the steering wheel 2, a steering torque Ts corresponding to the magnitude of the intervention operation is generated, and the value of the intervention coefficient α (and thus the follow-up command) according to the generated steering torque Ts. The tolerance range also changes.

At this time, limit the following command TC is not performed (restricted value u is directly output y, therefore the follow command TC) When, as indicated by a dotted line in the figure, the target following control, target angle theta ^* and In order to generate a large follow-up command TC (and thus autopilot torque) against an intervention operation that acts in the direction of increasing the deviation of the driver, the driver needs to operate the steering wheel with a very large force for overriding .

  On the other hand, in this embodiment, as the steering torque Ts increases, the intervention coefficient α approaches zero, and the allowable range of the follow-up command TC becomes narrower. When the absolute value | u | of the restriction target value increases beyond the allowable range, the follow-up command TC is limited to the allowable guard value ± LM. Further, when | Ts |> B, the intervention coefficient α and consequently the follow-up command TC become zero, and only assist control is performed. Therefore, the driver can operate the steering wheel without requiring extra force. it can.

<Effect>
As described above, in the electric steering system 1, when an intervention operation by the driver is detected, the follow-up command TC is decreased according to the magnitude (here, the magnitude of the steering torque Ts), whereby the assist command AC and Interference is suppressed. For this reason, when the driver performs an intervention operation, the target tracking control can be shifted to the assist control without causing the driver to feel uncomfortable.

  The electric steering system 1 does not directly limit the follow-up command TC generated by the target follow-up control calculation unit 30, but instead uses an internal value used in the calculation in the target follow-up control calculation unit 30, here, in the case of integral calculation. The follow-up command TC is restricted by restricting the value accumulated in the value so as not to exceed the guard value LM. For this reason, it is possible to suppress the occurrence of windup in the integral calculation, and it is possible to realize a smooth and uncomfortable intervention.

  Further, since the follow-up command TC does not become a value larger than the guard value LM and there is no possibility of overflow when the processing is realized by software, a highly safe software structure can be constructed.

<Modification>
In the present embodiment, the limit calculator 34 obtains the guard value LM from the intervention coefficient α, and if the limit target value u is within the allowable range −LM to LM, the limit target value u is left as it is, and the allowable range −LM to LM. If it is outside, the output y is obtained by limiting the restriction target value u to the guard value ± LM. The calculation in the limit calculator 34 is not limited to this. For example, the output y may be simply obtained by multiplying the limit target value u by the intervention coefficient α.

  In this case, as shown in FIG. 7, when an intervention operation is detected and the intervention coefficient α is set to less than 1, the operation is the same as in the above embodiment except that the restriction of the follow-up command TC is immediately started. Therefore, the same effect as that of the above embodiment can be realized by a simpler calculation.

  In the present embodiment, the target follow-up control calculation unit 30 includes a characteristic determiner 32 and an integrator 33, and the characteristic determiner 32 is configured to give a gain that determines the control characteristic of PID control. However, it is not limited to this. For example, a phase lead / lag compensator may be used as the characteristic determiner 32, or a controller designed by H∞ control may be used.

[Second Embodiment]
A second embodiment will be described.
Since the basic configuration of this embodiment is the same as that of the first embodiment, the description of the common configuration will be omitted, and the description will focus on the differences.

In the first embodiment, the limit calculator 34 is arranged so that the output of the adder constituting the integrator 33 becomes the limit target value u and the output y becomes the follow-up command TC.
On the other hand, in this embodiment, as shown in FIG. 8, the limit calculator 34 has the previous value TC [n-1] of the follow-up command as the limit target value u, and the output y is integrated by the adder. It arrange | positions so that it may become the addition value added to the value TM. Further, not only the intervention coefficient α but also the integration target value TM is input to the limit calculator 34.

  Then, as shown in FIG. 9, the limit calculator 34 first reads the target value u (the previous value TC [n−1] of the follow-up command), the intervention coefficient α, and the integration target value TM (S310). As in the first embodiment, the guard value LM is calculated by multiplying the upper limit value CL by the intervention coefficient α (S320).

  Next, it is determined whether or not the absolute value | u + TM | of the addition value of the restriction target value u and the integration target value TM is equal to or greater than the guard value LM (S330). That is, if the calculation is performed without any limitation, it is determined whether or not the addition result of the adder constituting the integrator 33 exceeds the allowable range −LM to LM of the follow-up command TC.

When u + TM is a value within the allowable range (| u + TM | <LM) (S330-NO), the restriction target value u is set as it is as output y without being restricted (S340).
If u + TM is a value outside the allowable range (| u + TM | ≧ LM) (S330-YES), it is determined whether u + TM is a non-negative value (S350).

  When u + TM is a non-negative value (S350-YES), the result obtained by subtracting the integration target value TM from the guard value (the upper limit value of the allowable range) LM is set as the output y (S360). Thereby, the addition result in the adder constituting the integrator 33 is LM.

  When u + TM is a negative value (S350-NO), the result obtained by subtracting the integration target value TM from the negative guard value (lower limit of allowable range) -LM is set as the output y (S370). As a result, the addition result in the adder constituting the integrator 33 is −LM.

<Effect>
According to the present embodiment, it is possible to obtain the same operational effects as in the case of the first embodiment.
<Modification>
In the present embodiment, the limit calculator 34 obtains the guard value LM from the intervention coefficient α, and if the added value of the limit target value u and the integration target value TM is within the allowable range −LM to LM, the limit target value u. Is output y, and if it is outside the allowable range −LM to LM, the output y is limited so that the addition result of the adder constituting the integrator 33 becomes the guard value ± LM. The calculation in the limit calculator 34 is not limited to this. For example, the output y may be simply obtained by multiplying the limit target value u by the intervention coefficient α.

[Third Embodiment]
A third embodiment will be described.
Since the basic configuration of this embodiment is the same as that of the first embodiment, the description of the common configuration will be omitted, and the description will focus on the differences.

  Unlike the first embodiment, in this embodiment, as shown in FIG. 10, the limit calculator 34 is an adder that forms the integrator 33 with the output y of the characteristic determiner 32 as the limit target value u. Are arranged so as to be the integration target value TM supplied to. Further, not only the intervention coefficient α but also the previous value TC [n−1] of the follow-up command is input to the limit calculator 34.

  Then, as shown in FIG. 11, the limit calculator 34 first reads the limit target value u (output of the characteristic determiner 32), the intervention coefficient α, and the previous value TC [n−1] of the follow-up command (S410). As in the first embodiment, the guard value LM is calculated by multiplying the upper limit value CL by the intervention coefficient α (S420).

  Next, it is determined whether or not the absolute value | u + TC [n−1] | of the addition value between the restriction target value u and the previous value TC [n−1] of the follow-up command is equal to or greater than the guard value LM (S430). . That is, if the calculation is performed without adding a restriction, it is determined whether or not the addition result of the adder constituting the integrator 33 exceeds the allowable range −LM to LM of the follow-up command TC.

  When u + TC [n−1] is a value within the allowable range (| u + TC [n−1] | <LM) (S430-NO), the restriction target value u is set as it is as the output y without being restricted (S440). ).

  When u + TC [n−1] is a value outside the allowable range (| u + TC [n−1] | ≧ LM) (S430—YES), it is determined whether u + TC [n−1] is a non-negative value. (S450).

  When u + TC [n-1] is a non-negative value (S450-YES), the result obtained by subtracting the previous value TC [n-1] of the follow-up command from the guard value (the upper limit value of the allowable range) LM is set as the output y. (S460). Thereby, the addition result in the adder constituting the integrator 33 is LM.

  When u + TC [n-1] is a negative value (S450-NO), the result obtained by subtracting the previous value TC [n-1] of the follow-up command from the negative guard value (lower limit value of the allowable range) -LM is output. Set as y (S470). As a result, the addition result in the adder constituting the integrator 33 is −LM.

<Effect>
According to the present embodiment, it is possible to obtain the same operational effects as in the case of the first embodiment.
Further, in this embodiment, since the restriction is imposed on the upstream side as compared with the above embodiment, when there is another application using the output (here, the integration target value TM), interference with the application is prevented. Can be suppressed. Because of such an interference polarity effect, the versatility of the software architecture can be improved.

[Fourth Embodiment]
A fourth embodiment will be described.
Since the basic configuration of this embodiment is the same as that of the first embodiment, the description of the common configuration will be omitted, and the description will focus on the differences.

  Unlike the first embodiment, in this embodiment, as shown in FIG. 12, the limit calculator 34 is configured so that the output of the subtractor 31 becomes the limit target value u and the output y becomes the input of the characteristic determiner 32. Is arranged. Further, not only the intervention coefficient α but also the previous value TC [n−1] of the follow-up command is input to the limit calculator 34.

  Then, as shown in FIG. 13, the restriction calculator 34 first reads the restriction target value u (output of the characteristic determiner 32), the intervention coefficient α, and the previous value TC [n−1] of the follow-up command (S510). ), The guard value LM is calculated by multiplying the upper limit CL by the intervention coefficient α, and the integral target value (characteristic determiner 32) by multiplying the restriction target value u by the gain K (fixed value) in the characteristic determiner 32. ) Is estimated (S520). Note that the actual gain K is not constant, but here the calculation is executed assuming that it is constant.

  Next, it is determined whether or not the absolute value | eTM + TC [n−1] | of the estimated value eTM and the previous value TC [n−1] of the follow-up command is equal to or greater than the guard value LM (S530). That is, when the limit calculator 34 does not apply a limit, it is determined whether or not the addition result of the adder constituting the integrator 33 exceeds the allowable range −LM to LM of the follow-up command TC.

  When eTM + TC [n-1] is a value within the allowable range (| eTM + TC [n-1] | <LM) (S530-NO), the restriction target value u is set as the output y without being restricted (S540). ).

  When eTM + TC [n-1] is a value outside the allowable range (| eTM + TC [n-1] | ≧ LM) (S530-YES), it is determined whether eTM + TC [n-1] is a non-negative value. (S550).

  When eTM + TC [n−1] is a non-negative value (S550−YES), the value obtained by subtracting the previous value TC [n−1] of the follow-up command from the guard value (the upper limit value of the allowable range) LM is divided by the gain K. The result is set as output y (S560). As a result, the addition result of the adder constituting the integrator 33 becomes substantially the same as that of the LM.

  When eTM + TC [n-1] is a negative value (S550-NO), gain is obtained by subtracting the previous value TC [n-1] of the follow-up command from the negative guard value (lower limit of the allowable range) -LM. The result divided by K is set as the output y (S570). As a result, the addition result of the adder constituting the integrator 33 is approximately the same as −LM.

<Effect>
According to the present embodiment, it is possible to obtain the same operational effects as in the case of the first embodiment.
Further, in the present embodiment, in order to limit on the upstream side as compared with the above embodiment, when there is another application using the output (here, the deviation between the target angle θ ^* and the actual angle θ), Interference with the application can be suppressed. Because of such an interference suppression effect, the versatility of the software architecture can be improved.

[Fifth Embodiment]
A fifth embodiment will be described.
Since the basic configuration of this embodiment is the same as that of the first embodiment, the description of the common configuration will be omitted, and the description will focus on the differences.

Unlike the first embodiment, in the present embodiment, as shown in FIG. 14, the limit calculator 34 has the target angle θ ^* and the actual angle θ become the target values u1 and u2 and the outputs y1 and y2 are subtracters. 31 inputs are arranged. Further, not only the intervention coefficient α but also the previous value TC [n−1] of the follow-up command is input to the limit calculator 34.

In the limit calculator 34, as shown in FIG. 15, first, target values u1, u2 (target angle θ ^* , actual angle θ), intervention coefficient α, previous value TC [n−1] of the follow-up command) (S610), the guard value LM is calculated by multiplying the upper limit value CL by the intervention coefficient α, and the gain K (fixed value) in the characteristic determiner 32 is obtained by subtracting the restriction target value u2 from the restriction target value u1. ) To calculate the estimated value eTM of the integration target value (output of the characteristic determiner 32) (S620). Note that the actual gain K is not constant, but here the calculation is executed assuming that it is constant.

  Next, it is determined whether or not the absolute value | eTM + TC [n−1] | of the estimated value eTM and the previous value TC [n−1] of the follow-up command is equal to or greater than the guard value LM (S630). That is, when the limit calculator 34 does not apply a limit, it is determined whether or not the addition result of the adder constituting the integrator 33 exceeds the allowable range −LM to LM of the follow-up command TC.

  When eTM + TC [n−1] is a value within an allowable range (| eTM + TC [n−1] | <LM) (S630−NO), the output values y1 and y2 are directly output without limiting the restriction target values u1 and u2. It sets (S640).

  When eTM + TC [n-1] is a value outside the allowable range (| eTM + TC [n-1] | ≧ LM) (S630-YES), it is determined whether eTM + TC [n-1] is a non-negative value. (S650).

  When eTM + TC [n−1] is a non-negative value (S650−YES), the value obtained by subtracting the previous value TC [n−1] of the follow-up command from the guard value (the upper limit value of the allowable range) LM is divided by the gain K. A value obtained by adding the restriction target value u2 to the result is set as the output y1, and the restriction target value u2 is set as the output y2 as it is (S660). As a result, the addition result of the adder constituting the integrator 33 becomes substantially the same as that of the LM.

  When eTM + TC [n-1] is a negative value (S650-NO), the gain obtained by subtracting the previous value TC [n-1] of the follow-up command from the negative guard value (lower limit of the allowable range) -LM A value obtained by adding the restriction target value u2 to the result divided by K is set as the output y1, and the restriction target value u2 is set as the output y2 as it is (S670). As a result, the addition result of the adder constituting the integrator 33 is approximately the same as −LM.

<Effect>
According to the present embodiment, it is possible to obtain the same operational effects as in the case of the fourth embodiment.
In the present embodiment, the restriction target value u1 is restricted in S660 and S670, but the restriction target value u2 may be restricted. In this case, y1 ← u1, y2 ← u1- (LM-TC [n−1]) / K is set in S660, and y1 ← u1, y2 ← u1-(− LM-TC [n−1]) is set in S670. / K

[Sixth Embodiment]
A sixth embodiment will be described.
Since the basic configuration of this embodiment is the same as that of the first embodiment, the description of the common configuration will be omitted, and the description will focus on the differences.

Unlike the first embodiment, in this embodiment, two limiting calculators 34A and 34B are provided before and after the integrator 33 as shown in FIG.
The limit calculator (hereinafter referred to as “preliminary limit calculator”) 34A is arranged so that the output of the characteristic determiner 32 becomes the limit target value u1, and the output y1 becomes the integration target value TM (input of the integrator 33). The intervention coefficient α and the follow-up command TC [n] are input.

  The limit calculator (hereinafter referred to as “post-limit calculator”) 34B is arranged so that the output of the integrator 33 becomes the limit target value u2, the output y2 becomes the follow-up command TC [n], and the intervention coefficient α is input. Has been.

  Then, as shown in FIG. 17, the pre-limit calculator 34A first reads the limit target value u1 (output of the characteristic determiner 32), the intervention coefficient α, and the follow-up command TC [n] (S710), and sets the upper limit value. A guard value LM is calculated by multiplying CL by an intervention coefficient α (S720).

  Next, it is determined whether or not the absolute value | TC [n] | of the follow-up command is greater than or equal to the guard value LM (S730). That is, it is determined whether or not the follow-up command TC [n] exceeds the allowable range −LM to LM.

When TC [n] is a value within the allowable range (| TC [n] | <LM) (S730-NO), the restriction target value u1 is set as it is as the output y1 without being restricted (S740).
When TC [n] is a value outside the allowable range (| TC [n] | ≧ LM) (S730-YES), the output y1 is set to zero (S750).

That is, when the follow-up command TC is outside the allowable range, the limit value u1 (that is, the integration target value TM) is set to zero so that the accumulation of the internal value of the integrator 33 does not proceed.
On the other hand, as shown in FIG. 18, the post-limit computing unit 34B first reads the restriction target value u2 (output of the integrator 33), reads the intervention coefficient α (S810), and multiplies the upper limit value CL by the intervention coefficient α. To calculate the guard value LM (S820).

  Then, it is determined whether or not the absolute value | u2 | of the restriction target value is greater than or equal to the guard value LM (S830). That is, it is determined whether or not the integration result in the integrator 33 exceeds the allowable range −LM to LM.

If u2 is a value within the allowable range (| u2 | <LM) (S830-NO), the restriction target value u2 is set as the output y2 without being restricted (S840).
If u2 is a value outside the allowable range (| u2 | ≧ LM) (S830-YES), it is determined whether u2 is a non-negative value (S850).

When u2 is a non-negative value (S850-YES), the guard value (the upper limit value of the allowable range) LM is set as the output y2 (S860).
When u2 is a negative value (S850-NO), negative guard value (lower limit value of allowable range) -LM is set as output y2 (S870).

<Effect>
According to the present embodiment, it is possible to obtain the same operational effects as in the case of the first embodiment.
In the present embodiment, the limit calculators 34A and 34B perform the limit calculation at a plurality of locations, so that it is possible to more reliably limit an abnormal value or the like generated when the internal value is garbled.

  In the present embodiment, the pre-limit calculator 34A is arranged at a place where the output of the characteristic determiner 32 (input of the integrator 33) is restricted, but instead of this place, a point P1 or a point in the figure is used. You may arrange | position to P2. Specifically, when the front limit calculator 34A is arranged at the point P1, the output of the subtractor 31 becomes the limit target value u1, and the output y1 becomes the input of the characteristic determiner 32. In this case, the processing content in the front limit calculator 34A is the same as that shown in FIG.

On the other hand, when the front limit computing unit 34A is arranged at the point P2, the target angle θ ^* and the actual angle θ become the restriction target values u11 and u12, and the outputs y11 and y12 become the inputs to the subtractor 31. In this case, the processing content in the pre-limit calculator 34A may be changed so that y11 ← u11, y12 ← u12 in S740, and y11 = y12 in S750.

[Seventh Embodiment]
A seventh embodiment will be described.
Since the basic configuration of this embodiment is the same as that of the first embodiment, the description of the common configuration will be omitted, and the description will focus on the differences.

In the first embodiment, the target follow-up control calculation unit 30 uses a control structure in which a PID gain is given to the deviation between the target angle θ ^* and the actual angle θ by the characteristic determiner 32 and then integration is performed by the integrator 33. doing.

On the other hand, as shown in FIG. 19, the target follow-up control calculation unit 30a of the present embodiment has a subtractor 31 that obtains the deviation of the actual angle with respect to the target angle θ ^* and a follow-up command TC from the deviation obtained by the subtractor 31. A proportional component calculator 35 for calculating the proportional component of the following command, an integral component calculator 36 for calculating the integral component of the follow-up command TC from the deviation obtained by the subtractor 31, and a differentiation of the follow-up command TC from the deviation obtained by the subtractor 31. A differential component computing unit 37 that computes components, an adder 38 that obtains a follow-up command TC by adding the computation results of the computing units 35 to 37, and a target follow-up control computation unit 30 according to the intervention coefficient α. A limiting calculator 39 is provided for limiting the follow-up command TC by limiting the internal value to be used. Note that each of the calculators 35 to 37 is well-known, but particularly for the integral component calculator 36, a mathematical expression representing general integration control is bilinearly converted for discretization, and the conversion is performed. What implement | achieves the control structure based on the obtained numerical formula is used.

  The limit calculator 39 limits the output of the proportional component calculator 35 to the limit target value u1, the output of the differential component calculator 37 to the limit target value u3, and the integral component calculator 36 limits the output of the adder that executes the integral calculation. A target value u2 is connected so that these outputs y1, y2, and y3 are input to the adder 38.

  Then, as shown in FIG. 20, the limit calculator 39 first reads the target values u1 to u3 and the intervention coefficient α (S910), and calculates the guard value LM by multiplying the upper limit value CL by the intervention coefficient α. (S920).

Next, it is determined whether or not the absolute value | u2 | of the restriction target value u2 (internal value of the integral calculation) is equal to or greater than the guard value LM (S930).
If | u | <LM (S930-NO), the restriction target values u1 to u3 are set as outputs y1 to y3 without being restricted (S940).

If | u | ≧ LM (S930-YES), it is determined whether the restriction target value u2 is a non-negative value (S950).
When the restriction target value u2 is a non-negative value (S950-YES), the values obtained by multiplying the restriction target values u1, u3 by the intervention coefficient α are output y1, y3, and the guard value (the upper limit value of the allowable range) LM is output. Set as y3 (S960).

  When the restriction target value u2 is a negative value (S950-NO), the values obtained by multiplying the restriction target values u1, u3 by the intervention coefficient α are output y1, y3, and the negative guard value (lower limit value of the allowable range) − LM is set as output y3 (S970).

<Effect>
According to the present embodiment, it is possible to obtain the same operational effects as in the case of the first embodiment.
In the present embodiment, each of the proportional component, the integral component, and the differential component of the follow-up command TC is limited. Therefore, there is another application that uses these outputs (either the proportional component, the integral component, or the differential component). In this case, interference with the application can be suppressed. Because of such an interference suppression effect, the versatility of the software architecture can be improved.

[Eighth Embodiment]
An eighth embodiment will be described.
Since the basic configuration of this embodiment is the same as that of the first embodiment, the description of the common configuration will be omitted, and the description will focus on the differences.

In the first embodiment, the target follow-up control calculation unit 30 is configured to limit the follow-up command TC based on the intervention coefficient α representing the detection result in the intervention detection unit 40.
On the other hand, in the present embodiment, as shown in FIG. 21, two types of intervention coefficients α and β representing detection results in the intervention detection unit 40a are generated, and the target follow-up control calculation unit 30 of the target tracking control calculation unit 30 is generated according to the intervention coefficient α. The tracking command TC is limited by limiting the internal value, and at the same time, the assist command AC is limited by limiting the internal value of the assist control calculation unit 20a according to the intervention coefficient β.

The follow-up control calculation unit 30 may have any configuration described in the first to eighth embodiments.
<Intervention detection unit>
As shown in FIG. 22, the intervention detection unit 40a first reads the steering torque Ts (S110), and in accordance with the conversion table prepared in advance, the intervention coefficient α, which corresponds to the absolute value | Ts | of the read steering torque Ts. β is calculated (S122).

The intervention detection unit 40a supplies the intervention coefficient α to the target follow-up control calculation unit 30, and supplies the intervention coefficient β to the assist control calculation unit 20a.
Here, since the conversion table used for calculating the intervention coefficient α is the same as that described in the intervention detection unit 40, description thereof is omitted. On the other hand, the conversion table used for calculating the intervention coefficient β outputs β = 0 when | Ts | ≦ A, β = 1 when | Ts | ≧ B, and | Ts when A <| Ts | <B. As | increases, a value that monotonously increases in the range of β = 0 to β = 1 is set. That is, the intervention detection unit 40a generates an intervention coefficient β that increases as the absolute value | Ts | of the steering torque increases, that is, as the degree of steering intervention increases. The meanings of the values A and B are the same as those described in the intervention detection unit 40.

  That is, the intervention coefficients α and β have a relationship of β = (1−α), and the intervention coefficient β is set to take a complementary value to the intervention coefficient α. Note that the relationship between the intervention coefficients α and β is not limited to this, and it is sufficient that the relationship α + β ≦ 1 is satisfied.

<Assist control calculation unit>
As shown in FIG. 23, the assist control calculation unit 20a is configured by the assist control calculation unit 20a according to the current command value calculation unit 21 having the same function as the assist control calculation unit 20 in the first embodiment and the intervention coefficient β. A limiting calculator 22 is provided for limiting the assist command AC by limiting internal values related to the calculation.

The limit calculator 22 is connected so that the output of the current command value calculator 21 is the limit target value u and its output y is the assist command AC.
Then, as shown in FIG. 24, the limit calculator 22 first reads the limit target value u (output of the current command value calculator 21), reads the intervention coefficient β (S1010), and sets a preset upper limit value of the assist command. A guard value LM is calculated by multiplying AL by an intervention coefficient β (S1020). For example, the upper limit value AL is set to the rated current of the motor.

  Then, it is determined whether or not the restriction target value u is greater than or equal to the guard value LM (S1030). That is, it is determined whether or not the calculation result of the current command value calculator 21 exceeds the allowable range 0 to LM.

When u <LM (S1030-NO), the restriction target value u is set as it is as output y without being restricted (S1040).
If u2 ≧ LM (S1030-YES), the guard value LM is set as the output y (S1050).

<Effect>
In this embodiment, when an intervention operation by the driver is detected, the follow-up command TC is decreased and the assist command AC is increased according to the magnitude (here, the magnitude of the steering torque Ts). Thus, according to the present embodiment, since the limitation of the assist command AC and the tracking command TC is linked, when the driver performs an intervention operation, the assist control is performed from the target tracking control without giving the driver a sense of incongruity. Since the range of the drive command DC obtained by adding the assist command AC and the follow-up command TC is limited, the capability of the motor 6 can be maximized.

  In the present embodiment, since both the assist command AC and the follow-up command TC are limited, control such as giving priority to the follow-up control over the assist control can be realized. Such control is useful, for example, when steering regardless of the driver's will in order to avoid accidents such as emergency avoidance.

<Modification>
The configuration of the assist control calculation unit 20a is not limited to that described above. For example, as shown in FIG. 25, instead of the limit calculator 22 described above, a current limit mechanism 23 that limits the output of the current command value calculator 21 using a fixed guard value, and an output of the current limit mechanism 23 May be constituted by a multiplier 24 that outputs an assist command AC.

[Ninth Embodiment]
A ninth embodiment will be described.
Since the basic configuration of this embodiment is the same as that of the eighth embodiment, the description of the common configuration will be omitted, and the description will focus on the differences.

In the eighth embodiment, the assist control calculation unit 20a limits the assist command AC according to the intervention coefficient β generated by the intervention detection unit 40a.
On the other hand, in this embodiment, as shown in FIG. 26, the assist control calculation unit 20b includes a follow-up command TC generated by the target follow-up control calculation unit 30 in addition to the intervention coefficient β generated by the intervention detection unit 40a. As a result, the assist command AC is limited.

<Assist control calculation unit>
As shown in FIG. 27, the assist control calculation unit 20b assists by limiting the internal value related to the calculation in the assist control calculation unit 20b according to the current command value calculator 21, the intervention coefficient β and the follow-up command TC. And a limit calculator 25 that limits the command AC.

The limit calculator 25 is connected so that the output of the current command value calculator 21 is the limit target value u and the output y is the assist command AC.
Then, in the limit calculator 25, as shown in FIG. 28, first, the limit target value u (output of the current command value calculator 21), the intervention coefficient β, and the follow-up command TC are read (S1110). The guard value LM is calculated by subtracting the multiplication value of the follow-up command TC and the intervention coefficient β (S1120).

  It is determined whether or not the restriction target value u is greater than or equal to the guard value LM (S1130). If u <LM (S1130-NO), the restriction target value u is directly set as the output y without being restricted (S1140). ), If u ≧ LM (S1130—YES), the guard value LM is set as the output y (S1150).

<Effect>
According to this embodiment, the same effect as that of the eighth embodiment can be obtained.
[Tenth embodiment]
Since the basic configuration of this embodiment is the same as that of the eighth embodiment, the description of the common configuration will be omitted, and the description will focus on the differences.

In the eighth embodiment, the intervention detection unit 40a calculates the intervention coefficients α and β according to the steering torque Ts.
On the other hand, in this embodiment, as shown in FIG. 29, the intervention detection unit 40b generates the intervention coefficients α and β according to the risk D representing the determination result in the risk determination unit 17.

<Danger degree determination part>
The risk determination unit 17 is a travel control acquired from a known in-vehicle control system or in-vehicle sensor, such as a stop control execution request flag by a PCS (pre-crash system), rain determination by a rain sensor, brightness determination by an illuminance sensor, The risk level D is calculated based on the information related to safety. There are various methods for calculating the degree of risk D, and the calculation method is not limited here, and therefore a specific description of the calculation method is omitted. However, it is assumed here that the lower the safety, the greater the risk level D.

<Intervention detection unit>
As shown in FIG. 30, the intervention detection unit 40b first reads the risk level D (S112), and calculates the intervention coefficients α and β corresponding to the read risk level D according to the conversion table prepared in advance (S122). ).

Then, the intervention detection unit 40b supplies the intervention coefficient α to the target follow-up control calculation unit 30, and supplies the intervention coefficient β to the assist control calculation unit 20a.
Here, the conversion table used for calculating the intervention coefficient α has a characteristic in which the degree of risk D is used instead of the absolute value | Ts | of the steering torque, and the characteristics in which the intervention coefficients α and β are interchanged. Except for this, it is the same as that described in the intervention detection unit 40a. That is, the intervention detection unit 40b generates an intervention coefficient α that increases as the degree of risk D increases, and an intervention coefficient β that decreases as the degree of risk D increases.

<Operation>
In the electric steering system 1 having such an intervention detection unit 40b, when the steering is performed by the driver, if the risk D is D ≦ A, the intervention coefficients α and β are α = 1 and β = 0, and there is no system intervention, and only assist control is executed.

  If the risk level D is A <D <B, the intervention coefficients α and β are set to values between 0 and 1 according to the risk level D. However, the relationship of α + β ≦ 1 is maintained. That is, as the risk level D increases, the ratio of the automatic steering torque generated by the target follow-up control increases from the assist torque generated by the assist control, thereby increasing the rate at which the system intervenes in the driver's steering.

If the degree of risk D is D ≧ B, the intervention coefficients α and β are α = 0 and β = 1, so that it is difficult for the driver to intervene and only target tracking control is executed.
That is, the driver performs control that makes it difficult to perform steering that increases the degree of risk D.

<Effect>
According to the present embodiment, by interlinking the restrictions of the assist command AC and the follow-up command TC, it is possible to realize system intervention without a sense of incongruity, and the range of the drive command DC obtained by adding the assist command AC and the follow-up command TC Therefore, the capacity of the motor 6 can be maximized.

  Here, the case where the intervention detection unit 40b generates both the intervention coefficients α and β on the assumption of the configuration of the eighth embodiment has been described, but only the intervention coefficient α is assumed on the assumption of the configuration of the first to seventh embodiments. May be configured to generate.

[Eleventh embodiment]
Since the basic configuration of this embodiment is the same as that of the eighth embodiment, the description of the common configuration will be omitted, and the description will focus on the differences.

In the eighth embodiment, the intervention detection unit 40a calculates the intervention coefficients α and β according to the steering torque Ts.
On the other hand, in this embodiment, as shown in FIG. 31, the steering torque Ts, the risk D representing the determination result in the risk determination unit 17, and the priority P representing the setting content in the priority setting unit 18. Accordingly, the intervention detection unit 40c generates the intervention coefficients α and β.

The risk determination unit 17 is the same as that described in the tenth embodiment, and a description thereof will be omitted.
<Priority setting section>
The priority setting unit 18 sets a priority P indicating how much priority is given to manual steering (assist control) by the driver and automatic steering (target tracking control) by the system. The priority P may be a fixed value, but may be changed according to the change speed of the risk D or the motor speed ω. However, the priority P takes a value of 0 ≦ P ≦ 1, and the larger P is, the greater the priority is given to the driver.

<Intervention detection unit>
As shown in FIG. 32, the intervention detection unit 40c first reads the steering torque Ts, the risk level D, and the priority level P (S114).

Next, the driver side intervention coefficients α _D and β _D according to the steering torque Ts and the system side intervention coefficients α _S and β _S according to the risk degree D are calculated according to a conversion table prepared in advance (S124).

The conversion table used for calculating the driver side intervention coefficients α _D and β _{D is} the same as that used for calculating the intervention coefficients α and β in the eighth embodiment (see FIG. 22). _The conversion table used for calculating _{S 1} and β _{S is} the same as that used for calculating the intervention coefficients α and β in the tenth embodiment (see FIG. 30). That is, the absolute value of the steering torque | Ts | larger the, driver-side intervention coefficient alpha _D small value, the driver-side intervention factor beta _D becomes a large value, the greater the risk D, large system-side intervention factor alpha _S value, the system-side intervention factor beta _S becomes a small value.

Next, using the calculated driver side intervention coefficients α _D and β _D , the system side intervention coefficients α _S and β _S , and the priority P, the intervention coefficients α and β are calculated according to the equations (1) and (2) ( S130).
α = Pα _D + (1-P) α _S (1)
β = Pβ _D + (1-P) β _S (2)
Then, the intervention detection unit 40c supplies the intervention coefficient α to the target tracking control calculation unit 30, and supplies the intervention coefficient β to the assist control calculation unit 20a.

<Effect>
According to the present embodiment, the intervention by the driver and the intervention by the system are arbitrated using the priority P, so that the switching between the assist control and the target follow-up control can be realized seamlessly without giving the driver a sense of incongruity. it can.

[Twelfth embodiment]
Since the basic configuration of this embodiment is the same as that of the first embodiment, the description of the common configuration will be omitted, and the description will focus on the differences.

In the first embodiment, the intervention detection unit 40 calculates the intervention coefficient α from the steering torque Ts using a single conversion map.
On the other hand, in this embodiment, the conversion map to be used is switched at the time of driver intervention (target tracking control → assist control) and at the time of return to target tracking control (assist control → target tracking control).

<Intervention detection unit>
As shown in FIG. 33, the intervention detection unit 40 first reads the steering torque Ts (S110), and determines whether the control states of the assist control and the target tracking control are returning to the target tracking control. (S116).

  The determination as to whether or not the return to the target tracking control is in progress is made by, for example, the timing at which the sign of the change speed (differential value) of the steering torque Ts detected by the torque sensor 4 is reversed, or by the limit calculator 34. Timing when the restriction is changed to the state where restriction is not made (for example, S230 in FIG. 5, S330 in FIG. 9, S430 in FIG. 11, S530 in FIG. 13, S630 in FIG. 15, S630 in FIG. S830 and the timing at which the determination in S930 is changed from YES to NO in S830 and FIG. 20 are determined to be returning until a certain time (for example, about several seconds to several tens of seconds) elapses.

  If not returning to the target tracking control (S116-NO), an intervention coefficient α corresponding to the read steering torque Ts is calculated according to a normal conversion table prepared in advance (S120). The normal conversion table is the same as that used in the first embodiment (see FIG. 3).

  On the other hand, if returning to the target follow-up control (S116-YES), an intervention coefficient α corresponding to the read steering torque Ts is calculated according to a prepared conversion table for return (S126).

Then, the intervention detection unit 40 supplies the intervention coefficient α calculated in S120 or S126 to the target tracking control calculation unit 30.
Here, the conversion table for restoration is such that α = 1 when | Ts | = 0, α = 0 when | Ts | ≧ B, and the exponential function as | Ts | decreases when 0 <| Ts | <B. Therefore, α is set to increase (or inversely proportional).

  That is, the absolute value | Ts | of the steering torque changes in the decreasing direction when the assist control is dominant due to the driver intervention and the driver intervention is released to return to the target tracking control. At this time, by using the conversion table for restoration, the increase in the intervention coefficient α becomes more gradual than when the normal conversion table is used.

<Effect>
According to the present embodiment, since the return to the target tracking control with high responsiveness is performed slowly and the sudden automatic steering toward the target value of the target tracking control is suppressed, the driver feels uncomfortable. Switching from assist control to target follow-up control can be realized without this.

[Thirteenth embodiment]
Since the basic configuration of this embodiment is the same as that of the twelfth embodiment, the description of the common configuration will be omitted, and the description will focus on the differences.

  In the twelfth embodiment, the intervention detection unit 40 calculates the intervention coefficient α from the steering torque Ts using a return conversion map different from the normal conversion map when returning to the target tracking control.

  On the other hand, in this embodiment, instead of switching the conversion map, the same effect as the conversion map for restoration is realized by calculating using the conversion map and applying a low-pass filter to the intervention coefficient α. doing.

<Intervention detection unit>
As shown in FIG. 34, the intervention detection unit 40 first reads the steering torque Ts (S110), and calculates the intervention coefficient α according to the read steering torque Ts according to the conversion table prepared in advance (S120).

  It is determined whether or not the control states of the assist control and the target tracking control are returning to the target tracking control (S140). This determination is the same as the processing in S116 described in the twelfth embodiment.

If returning to the target tracking control is not in progress (S140-NO), the calculation result in S120 is output as it is as the intervention coefficient α.
If returning to the target tracking control (S140-YES), the result obtained by applying the low-pass filter to the calculation result in S120 is output as the intervention coefficient α (S150).

The low-pass filter sets the cutoff frequency to an average driver steering speed (for example, about 0.1 Hz).
<Effect>
According to this embodiment, since the number of conversion maps to be used can be reduced as compared with the twelfth embodiment, the same effect can be obtained with a simple configuration and a small calculation load.

  Note that the cut-off frequency of the low-pass filter may be changed according to the degree of risk D described in the tenth embodiment, for example. Specifically, the higher the degree of risk D, the higher the cutoff frequency. Furthermore, when the degree of risk D is the maximum, the low pass filter calculation may not be executed. In this case, it is possible to quickly return to the target steering control when it is difficult to avoid danger by the driver's steering.

[Other Embodiments]
As mentioned above, although embodiment of this invention was described, it cannot be overemphasized that this invention can take a various form, without being limited to the said embodiment.

  (1) In the above embodiment, the case of performing the lane keep control as the target follow-up control is illustrated, but is not limited to this. For example, the motor rotation angle, the steering rotation angle, the yaw rate sensor, the tire turning angle, and the target Automatic steering torque based on deviation from values, lateral displacement from target position obtained by camera, laser radar, millimeter wave radar, etc., deviation from target locus obtained by GPS, etc., deviation from curvature obtained by road shape Any control can be used.

  (2) In the above embodiment, the intervention by the driver is detected based on the steering torque (output of the torque sensor), but the present invention is not limited to this, and a known steering intervention detection (determination) method can be used. For example, detection and determination may be made based on a deviation between a target value and a detection value in the target tracking control, a combination of the deviation, a motor rotation angular velocity, a torque sensor output, or the like.

  (3) The functions of one constituent element may be distributed to a plurality of constituent elements, or the functions of a plurality of constituent elements may be integrated into one constituent element. Further, at least a part of the configuration of the above embodiment may be replaced with a known configuration having the same function. In addition, at least a part of the configuration of the above embodiment may be added to or replaced with the configuration of the other embodiment.

  DESCRIPTION OF SYMBOLS 1 ... Electric steering system 4 ... Torque sensor 6 ... Motor 11 ... Vehicle speed sensor 20, 20a, 20b ... Assist control calculating part 21 ... Current command value calculator 22, 25 ... Limit calculator 23 ... Current limiting mechanism 24 ... Multiplier 30 ... target tracking control calculation unit 31 ... subtractor 32 ... characteristic determiner 33 ... integrator 34, 34A, 34B, 39 ... limit calculator 35 ... proportional component calculator 36 ... integral component calculator 37 ... differential component calculator 38 ... Adder 40, 40a, 40b ... Intervention detection unit 50 ... Adder 60 ... Motor drive circuit 100 ... Steering system mechanism

Claims (4)

An assist control means (20 ) for generating an assist command for generating an assist torque for reducing a steering load in accordance with a detected value of the steering torque;
Follow-up control means (30) for obtaining a target value of a physical quantity related to steering and generating a follow-up command for generating an automatic steering torque for causing the detected value of the physical quantity to follow the target value;
Motor driving means (60) for driving the motor (6) for generating the assist torque and the automatic steering torque in accordance with an added value of the assist command and the follow-up command;
Intervention detecting means (40, 40a, 4 0c) for detecting the intervention of the follow-up control by the driver and,
With
The follow-up control means includes
Deviation calculating means (31) for calculating a deviation between the target value of the physical quantity and the detected value of the physical quantity;
A characteristic determining means (32) for determining a control characteristic using an output value of the deviation calculating means;
Integrating means (33) for integrating the output value of the characteristic determining means to generate the follow-up command;
Either the integrated value integrated by the integrating means or the previous value of the integrated value added to the output value of the characteristic determining means when integrating by the integrating means is detected by the intervention detecting means as an internal value. depending on the degree of intervention that is, the so that the ratio of the assist torque is changed with respect to the automatic steering torque, and limiting means for limiting the pre-Symbol Internal value (3 4),
With
The limiting unit sets an allowable range of the internal value according to the degree of intervention detected by the intervention detecting unit, and limits the internal value so that the internal value becomes a value within the allowable range. A motor control device. Before SL limiting means, so that the ratio of the assist torque relative to the more the automatic steering torque is large degree of intervention to be detected by the interventional detecting means (40) increases, and limits the internal value The motor control device according to claim 1. The intervention detecting means (40c), in addition to the intervention in the follow-up control by the driver, detected a intervention in the assist control by the system,
The limiting means increases the ratio of the automatic steering torque with respect to the assist torque as the degree of intervention in the follow-up control by the driver detected by the intervention detection means increases, and assists by the system detected by the intervention detection means. The motor control device according to claim 1, wherein the internal value is limited so that the ratio of the automatic steering torque to the assist torque increases as the degree of intervention in control increases. Said limiting means, the sum of the assist command and the following command is any of claims 1 to 3, characterized in that to limit to a value within the range permitted by the control of the motor The motor control apparatus of Claim 1.

Images