Control device and control method

Abstract

In the present invention, a controller includes: a reception unit that acquires sprung acceleration, which is acceleration in the vertical direction, roll rate, and pitch rate from an IMU included in a vehicle; a sprung acceleration calculation unit that calculates the sprung acceleration of each wheel of the vehicle on the basis of the values input to the reception unit and specifications of the vehicle; a stroke speed calculation AI that calculates the stroke speed of a suspension device of each wheel using a mathematical model of the vehicle that receives, as input, the sprung acceleration of each wheel and a value indicating the force that is generated by a corresponding variable damper; a sprung speed calculation AI that calculates the spring speed of each wheel; and a ride comfort control unit that determines a control amount for controlling the variable damper from the sprung speed and the stroke speed.

Description

Control device and control method

[0001] The present disclosure relates to a control device and a control method for controlling a force generating device of a shock absorber provided in a vehicle.

[0002] Patent Document 1 discloses a control device that provides three acceleration sensors to a vehicle and controls four dampers of the vehicle based on detection signals from these acceleration sensors. Patent Document 2 discloses a control device that extracts road surface input components from wheel speed fluctuations and controls dampers without providing dedicated sensors.

[0003] International Publication No. WO 2022 / 168683 International Publication No. WO 2014 / 002444

[0004] The control device disclosed in Patent Document 1 requires three dedicated acceleration sensors to control the dampers, resulting in high manufacturing costs. On the other hand, the control device disclosed in Patent Document 2 can control the dampers using signals from general-purpose wheel speed sensors. However, to extract the road surface input component from wheel speed fluctuations, the wheel speed sensors must detect the longitudinal movement of the wheels of the vehicle due to the stroke of the suspension device caused by the road surface input. Therefore, in situations where wheels with a low friction coefficient slip, such as on packed snow or icy roads, the wheel speed fluctuations due to slip are large, making it difficult to estimate the vehicle's vertical vibration from the wheel speeds.

[0005] An object of one embodiment of the present invention is to provide a control device and a control method that are capable of controlling a force generating device of a shock absorber without using a dedicated sensor.

[0006] A control device according to one embodiment of the present invention is a control device that controls a force generating device of a shock absorber provided on a vehicle, and includes: a motion state acquisition unit that acquires vertical acceleration, which is acceleration in the vertical direction, a roll rate, and a pitch rate from a multipurpose sensor provided on the vehicle; a sprung acceleration acquisition unit that calculates the sprung acceleration of each wheel of the vehicle based on values ​​input to the motion state acquisition unit and specifications of the vehicle; a stroke speed acquisition unit that calculates the stroke speed of the shock absorber of each wheel using a mathematical model of the vehicle that inputs the sprung acceleration of each wheel and values ​​indicating the force generated by the force generating device; a sprung speed acquisition unit that calculates the sprung speed of each wheel; and a control quantity determination unit that determines a control quantity for controlling the force generating device from the sprung speed and the stroke speed.

[0007] A control method according to one embodiment of the present invention is a control method for controlling a force generating device of a shock absorber provided on a vehicle, and includes: a motion state acquisition step of acquiring vertical acceleration, which is acceleration in the vertical direction, a roll rate, and a pitch rate from a device other than the shock absorber provided on the vehicle; a sprung acceleration acquisition step of calculating the sprung acceleration of each wheel of the vehicle based on the values ​​acquired in the motion state acquisition step and specifications of the vehicle; a stroke speed acquisition step of calculating the stroke speed of the shock absorber of each wheel using a mathematical model of the vehicle that inputs the sprung acceleration of each wheel and values ​​indicating the force generated by the force generating device; a sprung velocity acquisition step of calculating the sprung velocity of each wheel; and a control quantity determination step of determining a control quantity for controlling the force generating device from the sprung velocity and the stroke velocity.

[0008] A control device according to one embodiment of the present invention is a control device that controls a force generating device of a shock absorber provided on a vehicle, and includes: a motion state acquisition unit that acquires vertical acceleration, which is acceleration in the vertical direction, a roll rate, and a pitch rate from a multipurpose sensor provided on the vehicle; a sprung acceleration acquisition unit that determines the sprung acceleration of each wheel of the vehicle based on values ​​input to the motion state acquisition unit and specifications of the vehicle; a stroke acquisition unit that determines the stroke of the shock absorber of each wheel using a mathematical model of the vehicle that has as input the sprung acceleration of each wheel and values ​​indicating the force generated by the force generating device; a stroke speed acquisition unit that calculates the stroke of the shock absorber of each wheel and determines the stroke speed of the shock absorber of each wheel; a sprung velocity acquisition unit that determines the sprung velocity of each wheel using values ​​input to the motion state acquisition unit and the mathematical model of the vehicle; and a control quantity determination unit that determines a control quantity for controlling the force generating device from the sprung velocity and the stroke speed.

[0009] A control method according to one embodiment of the present invention is a control method for controlling a force generating device of a shock absorber provided on a vehicle, and includes: a motion state acquisition step of acquiring vertical acceleration, which is acceleration in the vertical direction, a roll rate, and a pitch rate from a device other than the shock absorber provided on the vehicle; a sprung acceleration acquisition step of determining a sprung acceleration of each wheel of the vehicle based on the values ​​acquired in the motion state acquisition step and specifications of the vehicle; a stroke acquisition step of determining a stroke of the shock absorber of each wheel using a mathematical model of the vehicle that has as input the sprung acceleration of each wheel and values ​​indicating the generated force of the force generating device; a stroke speed acquisition step of calculating a stroke speed of the shock absorber of each wheel by calculating the stroke of the shock absorber of each wheel; a sprung velocity acquisition step of calculating a sprung velocity of each wheel using the values ​​acquired in the motion state acquisition step and the mathematical model of the vehicle; and a control variable determination step of determining a control variable for controlling the force generating device from the sprung velocity and the stroke velocity.

[0010] According to one embodiment of the present invention, it is possible to control the force generating device of the shock absorber without using a dedicated sensor.

[0011] 10 is an explanatory diagram showing a vehicle to which a control device according to a first embodiment is applied. FIG. 11 is a block diagram showing a damper control unit in FIG. 1. FIG. 12 is an explanatory diagram showing an example of a damping force calculation unit. FIG. 13 is an explanatory diagram showing an example of a neural network for a stroke speed calculation AI and a sprung speed calculation AI. FIG. 14 is an explanatory diagram showing a learning method for a neural network for a sprung speed calculation AI. FIG. 15 is a characteristic line diagram showing frequency characteristics of gain and phase of a transfer function between vertical acceleration and an estimated sprung speed. FIG. 16 is a characteristic line diagram showing time variation of the sprung speed calculated by the sprung speed calculation AI. FIG. 17 is an explanatory diagram showing a vehicle to which a control device according to a second embodiment is applied. FIG. 18 is a block diagram showing the damper control unit in FIG. 1. FIG. 19 is an explanatory diagram showing a vehicle to which a control device according to a third or fourth embodiment is applied. FIG. 19 is a block diagram showing a damper control unit according to a third embodiment. FIG. 19 is a block diagram showing a damper control unit according to a fourth embodiment. FIG. 19 is an explanatory diagram showing a vehicle to which a control device according to a fifth embodiment is applied. FIG. 19 is a block diagram showing a damper control unit according to a fifth embodiment. FIG. 19 is a block diagram showing a damper control unit according to a modified embodiment. FIG. 19 is an explanatory diagram showing a vehicle to which a control device according to a sixth or seventh embodiment is applied. FIG. 19 is a block diagram showing a damper control unit according to a sixth or seventh embodiment. FIG. 13 is a block diagram showing a damper control unit according to an eighth embodiment.

[0012] Hereinafter, a control device according to an embodiment of the present invention will be described in detail with reference to the accompanying drawings, taking as an example a case where the control device is applied to a four-wheeled vehicle.

[0013] Fig. 1 shows a vehicle 1 to which the control device is applied. The vehicle 1 includes a suspension device 5, an IMU 9, and a controller 11. In Fig. 1, for example, left and right front wheels and left and right rear wheels (hereinafter collectively referred to as wheels 3) are provided on the underside of a vehicle body 2 that constitutes the body of the vehicle. The wheels 3 include tires 4, which act as springs that absorb small irregularities in the road surface.

[0014] The suspension device 5 is a shock absorber provided in the vehicle 1. The suspension device 5 is provided between the vehicle body 2 and the vehicle wheels 3. The suspension device 5 is composed of a suspension spring 6 (hereinafter referred to as the spring 6) and an adjustable damping force shock absorber (hereinafter referred to as the variable damper 7) provided in parallel with the spring 6 between the vehicle body 2 and the wheels 3.

[0015] 1 shows a case where one set of suspension devices 5 is provided between the vehicle body 2 and the wheels 3. However, a total of four sets of suspension devices 5 are provided individually and independently between, for example, four wheels 3 and the vehicle body 2, and only one of these sets is shown schematically in FIG.

[0016] Here, the variable damper 7 of the suspension device 5 is configured using a damping force adjustable hydraulic shock absorber interposed between the vehicle body 2 and the wheel 3. The variable damper 7 is a force generating device of the suspension device 5. The variable damper 7 is provided between the vehicle body 2 and the wheel 3, and constitutes a relative displacement suppression device that changes the force that suppresses the relative displacement between the vehicle body 2 and the wheel 3.

[0017] The variable damper 7 is provided with a variable damping force actuator 8, which is composed of a damping force adjustment valve or the like, for continuously adjusting the characteristics of the generated damping force (i.e., the damping force characteristics) from hard characteristics (hard characteristics) to soft characteristics (soft characteristics). Note that the variable damping force actuator 8 does not necessarily have to be configured to continuously adjust the damping force characteristics, and may be capable of adjusting the damping force in multiple stages, for example, two or more stages. Furthermore, the variable damper 7 may be of a pressure control type or a flow rate control type. The variable damper 7 may also be of a type that controls viscosity, such as a magnetorheological fluid or an electrorheological fluid.

[0018] The inertial measurement unit 9 (hereinafter referred to as IMU 9) is provided on the vehicle body 2, which is the so-called sprung part, and constitutes vehicle body behavior detection means for detecting the behavior of the vehicle body 2. The IMU 9 includes, for example, a triaxial angular velocity sensor and a triaxial acceleration sensor. The IMU 9 is a multipurpose sensor provided in the vehicle 1. Detection signals from the IMU 9 are used for, for example, traction control, braking control, anti-skid control, and automatic driving control of the vehicle 1.

[0019] The IMU 9 is attached to an arbitrary sensor position on the vehicle body 2. The IMU 9 detects the vertical sprung acceleration at the sensor position, the roll rate which is the angular velocity in the roll direction, and the pitch rate which is the angular velocity in the pitch direction, and outputs the detection signals to a controller 11 described below. By knowing the attachment position (sensor position) of the IMU 9 in advance, the controller 11 can grasp the vehicle motion. That is, if the vehicle body 2, which is the sprung mass, is considered a rigid body, the vertical sprung mass absolute velocity (sprung mass velocity) of each wheel can be geometrically calculated based on the sprung mass acceleration, roll rate, and pitch rate detected at an arbitrary point on the vehicle body 2. The sensor position of the IMU 9 may be, for example, the center of gravity of the vehicle body 2 or a position other than the center of gravity.

[0020] A CAN 10 (Controller Area Network) is connected to the controller 11 and also to various multi-purpose sensors of the vehicle, such as a wheel speed sensor, a vehicle speed sensor, a steering angle sensor, etc. The CAN 10 transmits various types of vehicle information including the rotation speed (wheel speed) of the wheels 3 and the vehicle speed, which is the speed of the vehicle. This allows the controller 11 to obtain information such as the wheel speed and vehicle speed through the CAN 10.

[0021] In this embodiment, the in-vehicle network will be described using a CAN as an example, but other in-vehicle networks such as CAN FD (CAN with Flexible Data Rate), FlexRay, or in-vehicle Ethernet may also be used.

[0022] The controller 11 is a control device that controls the variable damper 7 (force generating device). The controller 11 is configured, for example, by a microcomputer as a control device that controls the damping characteristics of the variable damper 7. The controller 11 has a storage unit 12 that includes a ROM, a RAM, a non-volatile memory, etc. The storage unit 12 of the controller 11 stores various programs, information (sensor positions, vehicle specification information), data, etc. for controlling the variable damper 7. The input side of the controller 11 is connected to the IMU 9 and is also connected to, for example, a CAN 10, which is a line network required for data communication. The output side of the controller 11 is connected to the variable damping force actuator 8 of the variable damper 7.

[0023] The controller 11 calculates (estimates) the stroke speed and sprung speed as vehicle state quantities based on, for example, data acquired from the IMU 9. At this time, the stroke speed is the relative speed between the sprung and unsprung parts, and is the piston speed of the variable damper 7. The controller 11 determines the force to be generated by the variable damper 7 (force generating device) of the suspension device 5 based on the estimated vehicle state quantities, and outputs a control signal (command current) to the variable damping force actuator 8 of the suspension device 5.

[0024] 1 and 2 , the controller 11 includes a damper control unit 13, a weight parameter storage unit 14, and a data reading unit 15. The damper control unit 13 includes a receiving unit 16, a vehicle state quantity calculation unit 17, and a ride comfort control unit 22. The weight parameter storage unit 14 may be a part of the storage unit 12 or may be separate from the storage unit 12.

[0025] The receiver 16 acquires data on sprung acceleration, roll rate, and pitch rate from the IMU 9. The receiver 16 is a motion state acquisition unit that acquires vertical acceleration (sprung acceleration), roll rate, and pitch rate from the IMU 9, which serves as a multi-purpose sensor provided in the vehicle 1. The receiver 16 also acquires data related to the vehicle's behavior (hereinafter referred to as behavior information) via the CAN 10. At this time, the behavior information includes, for example, longitudinal acceleration (longitudinal G), lateral acceleration (lateral G), steering angle, yaw rate, wheel speed, etc. Note that the controller 11 does not need to acquire data directly from the IMU 9, but may acquire the data via the CAN 10.

[0026] The vehicle state quantity calculation unit 17 calculates (estimates) stroke speed and sprung speed as vehicle state quantities based on data acquired from, for example, the IMU 9. The vehicle state quantity calculation unit 17 includes a sprung acceleration calculation unit 18, a stroke speed calculation AI 19, and a sprung speed calculation AI 20. Based on data acquired from the IMU 9, the vehicle state quantity calculation unit 17 estimates data related to suspension control required by the ride comfort control unit 22, specifically, instantaneous values ​​related to suspension control (sprung speed, stroke speed).

[0027] The sprung acceleration calculation unit 18 is a sprung acceleration acquisition unit that calculates the sprung acceleration of each wheel of the vehicle 1 based on the values ​​input to the receiving unit 16 and the specifications of the vehicle 1. The sprung acceleration calculation unit 18 calculates the vertical sprung acceleration of each wheel based on the data acquired by the receiving unit 16 from the IMU 9, the sensor position of the IMU 9, and vehicle specification information such as the shape, size, weight, wheelbase, and position of each wheel 3 of the vehicle body 2. Specifically, the sprung acceleration calculation unit 18 calculates the roll angular acceleration (roll rate) and pitch angular acceleration (pitch rate) at the sensor position from the data acquired from the IMU 9, and also calculates the vertical sprung acceleration of each wheel based on the sprung acceleration, roll angular acceleration, and pitch angular acceleration at the sensor position and the relationship between the sensor position and the position of each wheel 3 (tire position). The sprung acceleration calculation unit 18 constitutes an IMU signal conversion unit that converts the signal (IMU signal) output from the IMU 9 into the sprung acceleration at the tire position.

[0028] The stroke speed calculation AI 19 is a stroke speed acquisition unit that calculates the stroke speed of the suspension device 5 of each wheel using a mathematical model of the vehicle 1 that receives as input values ​​representing the sprung acceleration of each wheel and the force (damping force) generated by the variable damper 7. In this case, the stroke speed calculation AI 19 is an artificial intelligence model constructed by machine learning. In other words, the mathematical model of the vehicle 1 is an artificial intelligence (AI) model.

[0029] The stroke speed calculation AI 19 receives the sprung acceleration of each wheel output from the sprung acceleration calculation unit 18 and the damping force (F) generated by the variable damper 7 of each wheel output from the damping force calculation unit 21. The stroke speed calculation AI 19 calculates the stroke speed of the suspension device 5 of each wheel based on the sprung acceleration and the damping force. Specifically, the stroke speed calculation AI 19 receives the sprung acceleration of each wheel and the damping force of the variable damper 7 and estimates the stroke speed of each wheel using a neural network that has been trained to estimate the stroke speed of each wheel. The stroke speed calculation AI 19 reads out the weight parameters stored in the weight parameter storage unit 14 via the data reading unit 15. The stroke speed calculation AI 19 estimates the stroke speed based on the sprung acceleration, the damping force, and the weight parameters.

[0030] The sprung velocity calculation AI20 is a sprung velocity acquisition unit that calculates the sprung velocity of each wheel. In this case, the sprung velocity calculation AI20 is an artificial intelligence model constructed by machine learning. The sprung acceleration of each wheel output from the sprung acceleration calculation unit 18 is input to the sprung velocity calculation AI20, and the sprung velocity of each wheel of the vehicle is calculated. Specifically, the sprung acceleration of each wheel is input to the sprung velocity calculation AI20, and the sprung velocity calculation AI20 estimates the sprung velocity of each wheel using a neural network that has been trained to estimate the sprung velocity of each wheel of the vehicle. The sprung velocity calculation AI20 reads out the weight parameters stored in the weight parameter storage unit 14 via the data reading unit 15. The sprung velocity calculation AI20 estimates the sprung velocity based on the sprung acceleration and the weight parameters.

[0031] The weight parameters of the trained neural network are stored in the weight parameter storage unit 14. Therefore, the neural network of the stroke speed calculation AI 19 and the neural network of the sprung speed calculation AI 20 are configured using the weight parameters stored in the weight parameter storage unit 14.

[0032] The damping force calculation unit 21 is a damping force acquisition unit that acquires a damping force by inputting a control variable from the ride comfort control unit 22 as a value indicating the force generated by the variable damper 7 (force generating device), and feeds back this damping force to the stroke speed calculation AI 19. The damping force calculation unit 21 calculates a damping force (F) generated by the variable damper 7 based on the stroke speed (V21) output from the stroke speed calculation AI 19 and the command current value (i) output from the ride comfort control unit 22. As shown in FIG. 3 , the damping force calculation unit 21 variably sets the relationship between the stroke speed and the damping force in accordance with the command current value. The relationship between the stroke speed, the damping force, and the command current value is created, for example, based on test data created by the inventors. The damping force calculation unit 21 calculates (acquires) the damping force (F) generated by the variable damper 7 based on the stroke speed (V21) and the command current value (i), and outputs this damping force (F) to the stroke speed calculation AI 19.

[0033] The ride comfort control unit 22 is a control amount determination unit that determines a control amount for controlling the variable damper 7 based on the sprung speed and the stroke speed. The ride comfort control unit 22 calculates a control command value (command current value) for controlling the damping force of the variable damper 7 of the suspension device 5 based on instantaneous values ​​input from the stroke speed calculation AI 19 and the sprung speed calculation AI 20. At this time, the control amount for controlling the variable damper 7 is the control command value for controlling the damping force of the variable damper 7. Specifically, the ride comfort control unit 22 calculates a command current value as a control command value for improving the ride comfort of the vehicle based on, for example, bilinear optimal control, skyhook control, H∞ control, or the like. The ride comfort control unit 22 outputs a command current as a control signal to the damping force variable actuator 8 of the variable damper 7 based on the command current value. Note that the control amount determination unit may calculate a control command value for improving not only the ride comfort of the vehicle but also the handling stability.

[0034] Next, the configuration and learning method of the neural network of the stroke speed calculation AI 19 will be described with reference to FIGS.

[0035] The stroke speed calculation AI19 and the sprung mass velocity calculation AI20 are artificial intelligence (AI) and are configured using trained neural networks. Weight parameters, sprung mass acceleration, and damping force are input to the stroke speed calculation AI19. The neural network of the stroke speed calculation AI19 is configured using the weight parameters. The stroke speed calculation AI19 estimates and outputs the stroke speed of the suspension device 5 for each wheel by inputting the sprung mass acceleration and damping force into the neural network. As shown in FIG. 2 , the stroke speed calculation AI19 outputs the stroke speed as a vehicle state quantity to the ride comfort control unit 22 in the subsequent stage. The neural network of the stroke speed calculation AI19 is configured similarly to the neural network of the vehicle behavior estimation unit disclosed in, for example, Japanese Patent Application Laid-Open No. 2022-191913.

[0036] As an example, a specific configuration of the neural network of the stroke speed calculation AI 19 will be described with reference to FIG. 4 . The neural network of the stroke speed calculation AI 19 is a three-layer hierarchical neural network in which elements of an input layer (number of elements i) 201, a hidden layer (number of elements j) 202, and an output layer (number of elements k) 203 are hierarchically connected. Each element of the input layer 201 is connected to each element of the hidden layer 202 by a weight W1ij (i = 1 to I, j = 1 to J), and each element of the hidden layer 202 is connected to each element of the output layer 203 by a weight W2jk (j = 1 to J, k = 1 to K). Information on these weights (hereinafter referred to as weight parameters) is expressed as a determinant of the weights W1ij and W2jk. The weight parameters are calculated in advance through machine learning and stored in the weight parameter storage unit 14. Note that, while this example illustrates the simplest fully-connected neural network with a single hidden layer 202, this is not limiting. For example, it may be a neural network with two or more hidden layers 202.

[0037] The input layer 201 of the neural network of the stroke speed calculation AI 19 receives, for example, time-series data of sprung acceleration output from the sprung acceleration calculation unit 18, and also receives, for example, time-series data of damping force output from the damping force calculation unit 21. The output layer 203 outputs instantaneous values ​​of the stroke speed of the suspension device 5 of each wheel of the vehicle 1. The number of elements in the hidden layer 202 is generally determined from the number of elements in the input layer 201 and the output layer 203, but is set to a number that maximizes the accuracy of state estimation by the neural network. The number of elements in the output layer 203 is determined by the output specifications of the vehicle behavior estimation.

[0038] The neural network machine learning is performed based on data on vehicle state quantities (e.g., sprung speed, unsprung speed) and data on sprung acceleration, etc., previously acquired by a data acquisition vehicle. The data acquisition vehicle has the same specifications as the vehicle in which the vehicle state quantity calculation unit 17 (stroke speed calculation AI 19, sprung speed calculation AI 20) is installed, and is equipped with various sensors (acceleration sensors, etc.) that acquire vehicle state quantities. In the neural network machine learning, weight parameters are adjusted to learn the correlation between the vehicle state quantity data (sprung speed, unsprung speed) acquired by the data acquisition vehicle and data on sprung acceleration, etc. The weight parameters obtained as a result of learning are stored in the weight parameter memory unit 14.

[0039] Next, the configuration and learning method of the neural network of the sprung velocity calculation AI 20 will be described with reference to FIGS. 2, 4, 5 and 6.

[0040] The specific configuration of the sprung velocity calculation AI20 will be described. The sprung velocity calculation AI20 is an artificial intelligence (AI) and is configured by a trained neural network. Weight parameters and sprung acceleration are input to the sprung velocity calculation AI20. The neural network of the sprung velocity calculation AI20 is configured using the weight parameters. The sprung velocity calculation AI20 estimates and outputs the sprung velocity of each wheel by inputting the sprung acceleration to the neural network. As shown in FIG. 2, the sprung velocity calculation AI20 outputs the sprung velocity as a vehicle state quantity to the subsequent ride comfort control unit 22. The neural network of the sprung velocity calculation AI20 is configured similarly to the neural network of the vehicle behavior estimation unit disclosed in, for example, Japanese Patent Application Laid-Open No. 2022-191913. As shown in FIG. 4, the neural network of the sprung velocity calculation AI20 is configured similarly to the neural network of the stroke speed calculation AI19.

[0041] For example, time-series data of the sprung acceleration of each wheel is input to the input layer 201 of the neural network of the sprung velocity calculation AI 20. For example, instantaneous values ​​of the sprung velocity of the suspension device 5 assumed to be attached to each wheel of a vehicle are output to the output layer 203. The number of elements in the hidden layer 202 is generally determined from the number of elements in the input layer 201 and the output layer 203, but is set to a number that maximizes the accuracy of state estimation by the neural network. The number of elements in the output layer 203 is determined by the output specifications of the state estimation.

[0042] Figure 5 shows a learning method for the neural network of the sprung velocity calculation AI 20. Data created using the method shown in Figure 5 is provided as training data for the neural network machine learning. First, the sprung velocity and vertical acceleration (sprung acceleration) of each wheel are measured when the vehicle body vibrates on an actual vehicle. Then, the vertical acceleration is integrated on the frequency axis to calculate the sprung velocity. The calculated sprung velocity (estimated sprung velocity) is combined with the measured sprung acceleration to create training data. This training data is used to perform machine learning of the neural network. In the neural network machine learning, weight parameters are adjusted to learn the correlation between the calculated sprung velocity data and the sprung acceleration data. The weight parameters obtained as a result of learning are stored in the weight parameter memory unit 14.

[0043] The frequency axis integral characteristic is shown in Figure 6. As shown by the dotted line in Figure 6, in the real-time integral filter used in the prior art, in order to reduce the influence of gradients, etc., the gain of frequencies lower than the control band is reduced relative to the ideal integral characteristic. It can also be seen that this influence causes the phase to advance near the sprung resonance frequency where control is originally desired.

[0044] In contrast, in the frequency axis integral characteristics, the control band is the frequency axis integral range. In this case, the control band is a predetermined frequency band in which high-frequency noise or low-frequency phase shift is reduced. The control band is, for example, a band of approximately 0.5 to 5 Hz, and includes the sprung resonance frequency of approximately 1 to 2 Hz. As shown by the two-dot chain line in Figure 6, the frequency axis integral characteristics show a significant reduction in gain in the range outside the control band compared to the integral filter characteristics, while maintaining no phase shift. By using a neural network trained using such training data, it is possible to achieve both ride comfort on flat roads and situations where the sensor tilts, such as on gradients.

[0045] The neural network of the sprung velocity calculation AI 20 can improve the estimation accuracy for time series data by inputting measurement results for a certain period of time in the past. At this time, by varying the time width for the certain period of time in the past according to the frequency to be controlled (the cutoff frequency of the frequency axis integral), it is possible to achieve both computational resources and performance.

[0046] Next, the estimation result of the sprung velocity when traveling on a sloped road surface using the neural network of the sprung velocity calculation AI20 will be described with reference to Fig. 7. The solid line in Fig. 7 shows the time change of the sprung velocity (estimated sprung velocity) calculated by the neural network (sprung velocity calculation AI20) of the present invention. The dotted line in Fig. 7 shows the time change of the estimated sprung velocity by the neural network of the prior art.

[0047] As shown in Figure 7, the neural network of the prior art learns the sprung speed obtained by integral filtering as training data. Therefore, in the prior art, the estimated results change significantly when the gradient changes. In contrast, the neural network of the present invention (sprung speed calculation AI20) learns the sprung speed calculated by frequency axis integration. Therefore, it can be confirmed that the neural network of the present invention does not significantly change the estimated value even when the gradient changes.

[0048] In this embodiment, the stroke speed calculation AI19 and the sprung speed calculation AI20 perform machine learning to determine the correlation between the vehicle state quantity data acquired by the data acquisition vehicle and the data on the sprung acceleration, etc. However, the present invention is not limited to this. For example, a vehicle model corresponding to the vehicle on which the stroke speed calculation AI19 and the sprung speed calculation AI20 are installed may be constructed, and the stroke speed calculation AI19 and the sprung speed calculation AI20 may perform machine learning based on data acquired by a simulation using this vehicle model.

[0049] Thus, the controller 11 (control device) according to this embodiment has a receiving unit 16 (motion state acquisition unit) that acquires sprung acceleration (vertical acceleration), which is acceleration in the vertical direction, roll rate, and pitch rate from the IMU 9 (multipurpose sensor) possessed by the vehicle 1; a sprung acceleration calculation unit 18 (sprung acceleration acquisition unit) that calculates the sprung acceleration of each wheel of the vehicle 1 based on the values ​​input to the receiving unit 16 and the specifications of the vehicle 1; a stroke speed calculation AI 19 (stroke speed acquisition unit) that calculates the stroke speed of the suspension device 5 (buffer device) of each wheel using a mathematical model (AI) of the vehicle 1 that inputs the sprung acceleration of each wheel and values ​​indicating the force generated by the variable damper 7 (force generating device); a sprung speed calculation AI 20 (sprung speed acquisition unit) that calculates the sprung velocity of each wheel; and a ride comfort control unit 22 (control amount determination unit) that determines the control amount for controlling the variable damper 7 from the sprung velocity and stroke speed.

[0050] That is, in this embodiment, the control method controls the variable damper 7 (force generating device) of the suspension device 5 (buffer device) of the vehicle, and includes: a motion state acquisition step of acquiring sprung acceleration (vertical acceleration), which is acceleration in the vertical direction, a roll rate, and a pitch rate from a device (IMU 9) other than the suspension device 5 of the vehicle 1; a sprung acceleration acquisition step of determining the sprung acceleration of each wheel of the vehicle 1 based on the values ​​acquired in the motion state acquisition step and the specifications of the vehicle 1; a stroke speed acquisition step of determining the stroke speed of the suspension device 5 of each wheel using a mathematical model (AI) of the vehicle 1 that inputs the sprung acceleration of each wheel and values ​​indicating the force generated by the variable damper 7; a sprung velocity acquisition step of determining the sprung velocity of each wheel; and a control quantity determination step of determining a control quantity for controlling the variable damper 7 from the sprung velocity and the stroke velocity.

[0051] This makes it possible to improve the accuracy of estimating the stroke speed without using a sensor dedicated to suspension control, thereby improving cost performance. In other words, in this embodiment, by using a general-purpose sensor on the vehicle side such as the IMU 9, a dedicated sensor is not required, and cost advantages can be achieved.

[0052] Furthermore, in this embodiment, in preprocessing before inputting the IMU signal to the mathematical model (AI) of the vehicle 1, the IMU signal is converted to sprung acceleration in order to unify the signal dimensions and convert it into a value at the tire position. At this time, there is a high correlation between sprung acceleration and sprung velocity, and there is also a high correlation between sprung acceleration and stroke velocity. As a result, learning of the AI ​​(sprung velocity calculation AI 20, stroke velocity calculation AI 19) becomes easier, the AI ​​network size can be reduced, and although the load due to preprocessing increases, the AI ​​calculation scale can be reduced, resulting in a reduction in the overall calculation load of the controller 11.

[0053] That is, in this embodiment, AI technology is combined with the signal (IMU signal) from the IMU 9, which is expected to be installed as a standard in the vehicle 1, so that low system cost and high estimation accuracy can be achieved at the same time. In addition, the sprung acceleration calculation unit 18, which serves as an IMU signal conversion unit, unifies the signal dimension and inputs it to AI such as the stroke speed calculation AI 19. This also reduces the processing load of the AI.

[0054] In addition, the sprung velocity can be obtained by integrating the sprung acceleration. However, the detection signal of the sprung acceleration measured by the sensor is superimposed with low-frequency noise such as vehicle acceleration / deceleration, turning, and gradient. For this reason, it is necessary to remove low-frequency noise from the detection signal of the sprung acceleration using a high-pass filter (HPF) before the integration process. In this case, since a transfer function is used in the high-pass filter, a positive phase shift occurs in principle, and the phase of the low-frequency band signal that passes the cutoff frequency is inevitably advanced. Therefore, there are accuracy issues with the sprung velocity obtained by integrating the sprung acceleration.

[0055] In contrast, the sprung velocity calculation AI20 of this embodiment is an artificial intelligence model (AI) constructed by machine learning. The sprung velocity calculation AI20 includes a neural network that learns the correlation between sprung acceleration and sprung velocity. The neural network of the sprung velocity calculation AI20 learns an expected value in which unnecessary frequency bands are removed by offline frequency integration. That is, the sprung velocity calculation AI20 uses an artificial intelligence model trained on data acquired through testing from which low-frequency noise and phase lead effects have been removed. Therefore, the sprung velocity calculation AI20 can calculate the sprung velocity while suppressing the effects of phase shift and gradient, which are difficult to achieve with filter processing. Therefore, the sprung velocity calculation AI20 can obtain a sprung velocity closer to the true value from the sprung acceleration. In this way, applying AI to the calculation of the sprung velocity makes it possible to achieve ideal frequency characteristics that are difficult to achieve with conventional filter processing, thereby improving ride comfort.

[0056] Furthermore, the state quantities of the sprung mass, which are affected by vehicle-specific specifications such as the moment arm and the sensor mounting position (sensor position) of the IMU 9, are taken into consideration when calculating the sprung mass acceleration of each wheel. Therefore, the sprung mass velocity calculation AI 20 does not require tuning (learning) for each vehicle type, which further reduces the unit cost.

[0057] The stroke speed calculation AI19 in this embodiment is an artificial intelligence model (AI) constructed by machine learning. In this case, the controller 11 in this embodiment is provided with the stroke speed calculation AI19 separately from the sprung speed calculation AI20. To calculate the stroke speed, state quantities that affect the unsprung part, such as the settings of the suspension spring constant and damping force, are essential. For this reason, it is necessary to set the stroke speed calculation AI19 for each characteristic of the suspension device 5. In this way, the stroke speed calculation AI19 is divided into more detailed parts than the sprung speed calculation AI20, and by making them separate AI blocks, the tuning man-hours can be reduced.

[0058] The controller 11 (control device) of this embodiment further includes a damping force calculation unit 21 (damping force acquisition unit) that acquires a damping force using a command current value (control amount) as an input as a value indicating the force generated by the variable damper 7 (force generating device) and feeds the damping force back to the stroke speed calculation AI 19 (stroke speed acquisition unit). Even with the same damper, the damping force changes depending on temperature conditions, etc. That is, the damping force of the damper is affected by viscosity changes due to temperature changes in the working fluid. In contrast, the stroke speed calculation AI 19 receives feedback input of the damping force of the variable damper 7 in addition to the sprung acceleration. This allows the stroke speed calculation AI 19 to estimate the stroke speed with higher accuracy.

[0059] Furthermore, Patent Document 2 discloses a control device that controls a damper using a signal from a general-purpose wheel speed sensor. However, in order to extract the road surface input component from wheel speed fluctuations, the wheel speed sensor must detect the longitudinal movement of the wheel due to the stroke of the suspension system caused by the road surface input. Therefore, the closer the stroke direction of the suspension system is to the vertical direction, the less the wheel movement component in the longitudinal direction of the vehicle. Therefore, the control method disclosed in Patent Document 2 cannot be applied to some vehicle models.

[0060] In contrast, the controller 11 according to this embodiment includes a receiver 16 that acquires sprung acceleration, which is acceleration in the vertical direction, a roll rate, and a pitch rate from the IMU 9 of the vehicle 1, and a sprung acceleration calculator 18 that calculates the sprung acceleration of each wheel of the vehicle 1 based on the values ​​input to the receiver 16 and the specifications of the vehicle 1. At this time, the sprung acceleration calculator 18 calculates the sprung acceleration at the tire position based on the values ​​of the sprung acceleration, roll rate, and pitch rate at the sensor position acquired by the receiver 16 and the specifications of the vehicle 1. Therefore, the sprung acceleration calculator 18 can calculate the sprung acceleration of each wheel regardless of the vehicle type. Therefore, the control method of this embodiment can be applied to any vehicle type.

[0061] 8 and 9 show a second embodiment. The second embodiment is characterized in that the sprung velocity acquisition unit is configured by an integrator that integrates the sprung acceleration. In the second embodiment, the same components as those in the first embodiment are denoted by the same reference numerals, and their description will be omitted.

[0062] 8 shows a vehicle 1 to which a controller 31 according to the second embodiment is applied. The controller 31 according to the second embodiment is a control device that controls the variable damper 7 (force generating device). The controller 31 is configured, for example, by a microcomputer as a control device that controls the damping characteristics of the variable damper 7. The controller 31 according to the second embodiment is configured similarly to the controller 11 according to the first embodiment. The controller 31 acquires various data from the IMU 9 and the CAN 10. The output side of the controller 31 is connected to the damping force variable actuator 8 of the variable damper 7.

[0063] The controller 31 has a storage unit 12 including a ROM, a RAM, a non-volatile memory, etc. The storage unit 12 of the controller 31 stores various programs, information (sensor positions, vehicle specification information), data, etc. for controlling the variable damper 7. The controller 31 estimates the sprung velocity and stroke velocity as vehicle state quantities based on data acquired from the IMU 9. The controller 31 determines the force to be generated by the variable damper 7 (force generating device) of the suspension device 5 based on the estimated vehicle state quantities, and outputs a control signal (command current) to the variable damping force actuator 8 of the suspension device 5.

[0064] 8 and 9, the controller 31 includes a damper control unit 32, a weight parameter storage unit 14, and a data reading unit 15. The damper control unit 32 includes a receiving unit 16, a vehicle state quantity calculation unit 33, and a ride comfort control unit 22.

[0065] The vehicle state quantity calculation unit 33 calculates (estimates) the stroke speed and the sprung speed as vehicle state quantities based on, for example, data acquired from the IMU 9. The vehicle state quantity calculation unit 33 includes the sprung acceleration calculation unit 18, a stroke speed calculation AI 34, and an integrator 35. Based on the data acquired from the IMU 9, the vehicle state quantity calculation unit 33 estimates data related to suspension control required by the ride comfort control unit 22, specifically, instantaneous values ​​related to suspension control (sprung speed, stroke speed).

[0066] The stroke speed calculation AI 34 is a stroke speed acquisition unit that calculates the stroke speed of the suspension device 5 of each wheel using a mathematical model of the vehicle 1, which inputs values ​​indicating the sprung acceleration of each wheel and the force (damping force) generated by the variable damper 7. In this case, the stroke speed calculation AI 34 is an artificial intelligence model constructed by machine learning. The stroke speed calculation AI 34 is configured in the same manner as the stroke speed calculation AI 19 according to the first embodiment.

[0067] The stroke speed calculation AI 34 receives the sprung acceleration of each wheel output from the sprung acceleration calculation unit 18 and the command current value (i) output from the ride comfort control unit 22. The stroke speed calculation AI 19 calculates the stroke speed of the suspension device 5 of each wheel based on the sprung acceleration and the command current value (i). Specifically, the stroke speed calculation AI 34 receives the sprung acceleration of each wheel and the command current value (i) for the variable damper 7, and estimates the stroke speed of each wheel using a neural network that has been trained to estimate the stroke speed of each wheel. The stroke speed calculation AI 34 reads out the weight parameters stored in the weight parameter storage unit 14 via the data reading unit 15. The stroke speed calculation AI 34 estimates the stroke speed based on the sprung acceleration, the command current value (i), and the weight parameters.

[0068] The integrator 35 is a sprung velocity acquisition unit that calculates the sprung velocity of each wheel. The integrator 35 receives the sprung acceleration of each wheel output from the sprung acceleration calculation unit 18 and calculates the sprung velocity of each wheel of the vehicle. Specifically, the integrator 35 integrates the sprung acceleration of each wheel 3 to calculate the sprung velocity of each wheel. The integrator 35 also includes a filter (not shown) for compensating for the phase of the output signal. Therefore, the integrator 35 integrates the sprung acceleration of each wheel 3, performs filtering, and calculates the sprung velocity of each wheel.

[0069] The stroke speed output by the stroke speed calculation AI 34 and the sprung mass speed output by the integrator 35 are input to the ride comfort control unit 22. The ride comfort control unit 22 calculates a control command value (command current value) for controlling the damping force of the variable damper 7 of the suspension device 5 based on the instantaneous values ​​of the stroke speed and the sprung mass speed.

[0070] Thus, the second embodiment can achieve substantially the same effects as the first embodiment. The stroke speed, including unsprung mass information, needs to be estimated based on sprung mass information, etc. Therefore, in the second embodiment, the stroke speed is calculated by the stroke speed calculation AI 34 equipped with AI. Meanwhile, sprung mass information (sprung mass acceleration, sprung mass velocity) can be obtained from data from the IMU 9 or the like or through various calculations. Therefore, in the second embodiment, the sprung mass velocity is calculated by integrating the sprung mass acceleration. As a result, the configuration of the controller 31 can be simplified, and manufacturing costs can be reduced.

[0071] 10 and 11 show a third embodiment. The third embodiment is characterized in that the stroke speed calculation AI calculates the stroke speeds of the suspension devices for four wheels based on the sprung accelerations of the four wheels and the forces (damping forces) generated by the variable dampers 7 for the four wheels. In the third embodiment, the same components as those in the first embodiment are designated by the same reference numerals, and their description will be omitted.

[0072] FIG. 10 shows a vehicle 1 to which a controller 41 according to a third embodiment is applied. The controller 41 according to the third embodiment is a control device that controls a variable damper 7 (force generating device). The controller 41 is configured, for example, by a microcomputer as a control device that controls the damping characteristics of the variable damper 7. The controller 41 according to the third embodiment is configured similarly to the controller 11 according to the first embodiment. The controller 41 acquires various data from the CAN 10. At this time, the data transmitted by the CAN 10 includes data output by various general-purpose sensors including, for example, an inertial measurement unit (IMU). The output side of the controller 41 is connected to the variable damping force actuator 8 of the variable damper 7.

[0073] The controller 41 has a storage unit 12 including a ROM, a RAM, a non-volatile memory, etc. The storage unit 12 of the controller 41 stores various programs, information (sensor positions, vehicle specification information), data, etc. for controlling the variable damper 7. The controller 41 estimates the sprung velocity and stroke velocity as vehicle state quantities based on data acquired from the CAN 10. The controller 41 determines the force to be generated by the variable damper 7 (force generating device) of the suspension device 5 based on the estimated vehicle state quantities, and outputs a control signal (command current) to the variable damping force actuator 8 of the suspension device 5.

[0074] 10 and 11 , the controller 41 includes a damper control unit 42, a weight parameter storage unit 14, and a data reading unit 15. The damper control unit 42 includes a receiving unit 16, a vehicle state quantity calculation unit 43, and a control quantity determination unit 48.

[0075] The vehicle state quantity calculation unit 43 calculates (estimates) stroke speed and sprung speed as vehicle state quantities based on data acquired from, for example, the CAN 10. The vehicle state quantity calculation unit 43 includes a sprung acceleration calculation unit 44, a stroke speed calculation AI 45, and a sprung speed calculation AI 46. Based on data acquired from the CAN 10, the vehicle state quantity calculation unit 43 estimates data related to suspension control required by the ride comfort control unit 49, specifically, instantaneous values ​​related to suspension control (sprung speed, stroke speed).

[0076] The sprung acceleration calculation unit 44 is a sprung acceleration acquisition unit that calculates the sprung acceleration for each of the four wheels of the vehicle 1 based on the values ​​input to the receiving unit 16 and the specifications of the vehicle 1. Specifically, the sprung acceleration calculation unit 44 calculates the vertical sprung acceleration for each of the four wheels based on the IMU signal data acquired by the receiving unit 16 from the CAN 10, the sensor position of the IMU, and vehicle specification information such as the shape, size, weight, wheelbase, and position of each wheel 3 of the vehicle body 2. At this time, the IMU signal data includes data on the vertical acceleration (vertical acceleration), roll rate, and pitch rate at the sensor position. The sprung acceleration calculation unit 44 constitutes an IMU signal conversion unit that converts the signal (IMU signal) output from the IMU into the sprung acceleration at the tire position.

[0077] The stroke speed calculation AI 45 is a stroke speed acquisition unit that calculates the stroke speeds of the suspension devices 5 for four wheels using a mathematical model of the vehicle 1 that receives as input values ​​representing the sprung accelerations for four wheels and the forces (damping forces) generated by the variable dampers 7 for four wheels. Therefore, a total of eight signals representing the sprung accelerations for four wheels and the damping forces of the variable dampers 7 for four wheels are input to the stroke speed calculation AI 45, and the stroke speeds of the suspension devices 5 for four wheels are calculated from these eight signals. In this case, the stroke speed calculation AI 45 is an artificial intelligence model constructed by machine learning. The stroke speed calculation AI 45 includes a neural network similar to that of the stroke speed calculation AI 19 according to the first embodiment.

[0078] The stroke speed calculation AI 45 receives the sprung accelerations for the four wheels output from the sprung acceleration calculation unit 44 and the damping forces (F) generated by the variable dampers 7 for the four wheels output from the damping force calculation unit 47. The stroke speed calculation AI 45 calculates the stroke speeds of the suspension devices 5 for the four wheels based on the sprung accelerations and the damping forces. Specifically, the stroke speed calculation AI 45 receives the sprung accelerations for the four wheels and the damping forces of the variable dampers 7 for the four wheels and estimates the stroke speeds for the four wheels using a neural network that has been trained to estimate the stroke speeds for the four wheels. The stroke speed calculation AI 45 reads out the weight parameters stored in the weight parameter storage unit 14 via the data reading unit 15. The stroke speed calculation AI 45 estimates the stroke speeds for the four wheels based on the sprung accelerations for the four wheels, the damping forces for the four wheels, and the weight parameters.

[0079] The sprung velocity calculation AI46 is a sprung velocity acquisition unit that calculates the sprung velocity of each wheel. The sprung velocity calculation AI46 is an artificial intelligence model constructed by machine learning. The sprung accelerations for the four wheels output from the sprung acceleration calculation unit 44 are input to the sprung velocity calculation AI46, which calculates the sprung velocity of each wheel of the vehicle. Specifically, the sprung velocity calculation AI46 receives the sprung accelerations for the four wheels and estimates the sprung velocity of each wheel using a neural network that has been trained to estimate the sprung velocity of each wheel of the vehicle. The sprung velocity calculation AI46 reads the weight parameters stored in the weight parameter storage unit 14 via the data reading unit 15. The sprung velocity calculation AI46 estimates the sprung velocity based on the sprung acceleration and the weight parameters. The sprung velocity calculation AI46 includes a neural network similar to that of the sprung velocity calculation AI20 according to the first embodiment.

[0080] The damping force calculation unit 47 is a damping force acquisition unit that acquires damping forces by inputting the control amount from the control amount determination unit 48 as values ​​indicating the forces generated by the variable dampers 7 (force generating devices), and feeds back the acquired damping forces to the stroke speed calculation AI 19. The damping force calculation unit 47 calculates the damping forces (F) generated by the variable dampers 7 for the four wheels based on the stroke speeds (V21) for the four wheels output from the stroke speed calculation AI 45 and the command current values ​​(i) output from the control amount determination unit 48. The damping force calculation unit 47 is configured similarly to the damping force calculation unit 21 according to the first embodiment. The damping force calculation unit 47 variably sets the relationship between the stroke speeds and the damping forces in accordance with the command current values. The damping force calculation unit 47 calculates (acquires) the damping forces (F) generated by the variable dampers 7 for the four wheels based on the stroke speeds (V21) for the four wheels and the command current values ​​(i), and outputs these damping forces (F) to the stroke speed calculation AI 45.

[0081] The control amount determination unit 48 determines the control amount for controlling the variable damper 7 based on the sprung speed and the stroke speed. The control amount determination unit 48 includes a ride comfort control unit 49, a handling stability control unit 50, and a control command management unit 51.

[0082] The ride comfort control unit 49 determines the control amount for controlling the variable damper 7 from the sprung speed and the stroke speed. The ride comfort control unit 49 is configured in the same manner as the ride comfort control unit 22 according to the first embodiment. The ride comfort control unit 49 calculates a control command value (command current value) for controlling the damping force of the variable damper 7 of the suspension device 5 based on the instantaneous values ​​input from the stroke speed calculation AI 45 and the sprung speed calculation AI 46. Specifically, the ride comfort control unit 49 calculates the command current value as the control command value for improving the ride comfort of the vehicle based on, for example, bilinear optimal control, skyhook control, H∞ control, or the like.

[0083] The handling stability control unit 50 calculates a control command value (command current value) for improving the handling stability of the vehicle. The handling stability control unit 50 has, for example, anti-roll control, anti-divesquat control, etc. The handling stability control unit 50 receives inputs such as the steering angle and vehicle speed from, for example, the receiving unit 16. The handling stability control unit 50 calculates a command current value as a control command value for improving the handling stability of the vehicle based on the input values ​​such as the steering angle and vehicle speed.

[0084] The control command management unit 51 outputs a control command value (command current value) for varying (adjusting) the damping force of the variable damper 7 in accordance with the control command value from the ride comfort control unit 49 and the control command value from the driving stability control unit 50. The control command management unit 51 integrates the control command value from the ride comfort control unit 49 and the control command value from the driving stability control unit 50 to output a final control command value. The control command management unit 51 compares the control command value from the ride comfort control unit 49 with the control command value from the driving stability control unit 50, selects the larger control command value, and outputs it as the final control command value (command current value). Specifically, the control command management unit 51 selects the hard value from the control command value from the ride comfort control unit 49 and the control command value from the driving stability control unit 50. At this time, the control command management unit 51 performs the same processing for the control command values ​​of each of the four wheels. A command current is supplied as a control signal to the damping force variable actuator 8 of the variable damper 7 based on the control command value. As a result, the damping force of the variable damper 7 is controlled by the controller 41 .

[0085] Thus, the third embodiment can achieve substantially the same effects as the first embodiment. The stroke speed calculation AI 45 calculates the stroke speeds of the suspension devices 5 for four wheels based on the sprung accelerations of the four wheels and the forces (damping forces) generated by the variable dampers 7 for the four wheels. Therefore, the accuracy of estimating the stroke speed can be improved compared to when the stroke speed is calculated based on the sprung accelerations and damping forces of each wheel.

[0086] 10 and 12 show a fourth embodiment. The fourth embodiment is characterized in that the stroke speed calculation AI calculates the stroke speed of the suspension device based on the sprung acceleration and the control command value (command current value) for the variable damper. In the fourth embodiment, the same components as those in the third embodiment are designated by the same reference numerals, and their description will be omitted.

[0087] 10 shows a vehicle 1 to which a controller 61 according to the fourth embodiment is applied. The controller 61 according to the fourth embodiment is a control device that controls a variable damper 7 (force generating device). The controller 61 is configured, for example, by a microcomputer as a control device that controls the damping characteristics of the variable damper 7. The controller 61 according to the fourth embodiment is configured similarly to the controller 41 according to the third embodiment. The controller 61 acquires various data from the CAN 10. The output side of the controller 61 is connected to the damping force variable actuator 8 of the variable damper 7.

[0088] The controller 61 has a storage unit 12 including a ROM, a RAM, a non-volatile memory, etc. The storage unit 12 of the controller 61 stores various programs, information (sensor positions, vehicle specification information), data, etc. for controlling the variable damper 7. The controller 61 estimates the sprung velocity and stroke velocity as vehicle state quantities based on data acquired from the CAN 10. The controller 61 determines the force to be generated by the variable damper 7 (force generating device) of the suspension device 5 based on the estimated vehicle state quantities, and outputs a control signal (command current) to the variable damping force actuator 8 of the suspension device 5.

[0089] 10 and 12, the controller 61 includes a damper control unit 62, a weight parameter storage unit 14, and a data reading unit 15. The damper control unit 62 includes a receiving unit 16, a vehicle state quantity calculation unit 63, and a control quantity determination unit 48.

[0090] The vehicle state quantity calculation unit 63 calculates (estimates) stroke speed and sprung speed as vehicle state quantities based on data acquired from, for example, the CAN 10. The vehicle state quantity calculation unit 63 includes a sprung acceleration calculation unit 44, a stroke speed calculation AI 64, and a sprung speed calculation AI 46. Based on data acquired from the CAN 10, the vehicle state quantity calculation unit 63 estimates data related to suspension control required by the ride comfort control unit 49, specifically, instantaneous values ​​related to suspension control (sprung speed, stroke speed).

[0091] The stroke speed calculation AI 64 according to the fourth embodiment is configured similarly to the stroke speed calculation AI 45 according to the third embodiment. However, the stroke speed calculation AI 64 is a stroke speed acquisition unit that calculates the stroke speeds of the suspension devices 5 for four wheels using a mathematical model of the vehicle 1 that receives as input the sprung accelerations of four wheels and the control command values ​​(command current values) for the variable dampers 7 for four wheels. The stroke speed calculation AI 64 receives a total of eight signals, namely the sprung accelerations of four wheels and the command current values ​​(i) for four wheels, and calculates the stroke speeds of the suspension devices 5 for four wheels from these eight signals. In this case, the stroke speed calculation AI 64 is an artificial intelligence model constructed by machine learning. The stroke speed calculation AI 64 includes a neural network similar to that of the stroke speed calculation AI 34 according to the second embodiment.

[0092] Thus, the fourth embodiment can also achieve substantially the same effects as the third embodiment. In the fourth embodiment, the stroke speed calculation AI 64 calculates the stroke speed of the suspension unit 5 based on the sprung acceleration and the control command value (command current value) for the variable damper. Therefore, compared to the third embodiment, where the stroke speed of the suspension unit 5 is calculated based on the sprung acceleration and the damping force value of the variable damper, there is no need to calculate the damping force, so the configuration of the controller 61 can be simplified and manufacturing costs can be reduced.

[0093] 13 and 14 show a fifth embodiment. The fifth embodiment is characterized in that the sprung speed at any one location is calculated using AI, the unsprung speed of the front wheels is calculated using AI, and the stroke speeds of the four wheels are calculated based on the sprung speed at one location and the unsprung speed of the front wheels. In the fifth embodiment, the same components as those in the third embodiment are designated by the same reference numerals, and their description will be omitted.

[0094] FIG. 13 shows a vehicle 1 to which a controller 71 according to a fifth embodiment is applied. The controller 71 according to the fifth embodiment is a control device that controls a variable damper 7 (force generating device). The controller 71 is configured, for example, by a microcomputer as a control device that controls the damping characteristics of the variable damper 7. The controller 71 according to the fifth embodiment is configured similarly to the controller 41 according to the third embodiment. The controller 71 acquires various data from the CAN 10. At this time, the data transmitted by the CAN 10 includes data output by various general-purpose sensors including, for example, an inertial measurement unit (IMU). The output side of the controller 71 is connected to the variable damping force actuator 8 of the variable damper 7.

[0095] The controller 71 has a storage unit 12 including a ROM, a RAM, a non-volatile memory, etc. The storage unit 12 of the controller 71 stores various programs, information (sensor positions, vehicle specification information), data, etc. for controlling the variable damper 7. The controller 71 estimates the sprung velocity and stroke velocity as vehicle state quantities based on data acquired from the CAN 10. The controller 71 determines the force to be generated by the variable damper 7 (force generating device) of the suspension device 5 based on the estimated vehicle state quantities, and outputs a control signal (command current) to the variable damping force actuator 8 of the suspension device 5.

[0096] 13 and 14 , the controller 71 includes a damper control unit 72, a weight parameter storage unit 14, and a data reading unit 15. The damper control unit 72 includes a receiving unit 16, a vehicle state quantity calculation unit 73, and a control quantity determination unit 48.

[0097] The vehicle state quantity calculation unit 73 calculates (estimates) stroke speed and sprung speed as vehicle state quantities based on, for example, data acquired from the CAN 10. The vehicle state quantity calculation unit 73 includes a sprung speed calculation unit 74, a four-wheel sprung speed calculation unit 75, a front-wheel sprung acceleration calculation unit 76, a front-wheel unsprung speed calculation unit 77, and a four-wheel stroke speed calculation unit 78. Based on data acquired from the CAN 10, the vehicle state quantity calculation unit 73 estimates data related to suspension control required by the ride comfort control unit 49, specifically, instantaneous values ​​related to suspension control (sprung speed, stroke speed).

[0098] The sprung velocity calculation AI74 is an artificial intelligence model constructed by machine learning. The sprung velocity calculation AI74 receives the sprung acceleration (vertical acceleration) at the sensor position of the IMU acquired by the receiving unit 16 from the CAN 10 and calculates the sprung velocity at the sensor position. Specifically, the sprung velocity calculation AI74 receives the sprung acceleration at the sensor position and estimates the sprung velocity at the sensor position using a neural network that has been trained to estimate the sprung velocity at the sensor position. The sprung velocity calculation AI74 reads out the weight parameters stored in the weight parameter storage unit 14 via the data reading unit 15. The sprung velocity calculation AI74 estimates the sprung velocity at the sensor position based on the sprung acceleration at the sensor position and the weight parameters.

[0099] The sprung velocity calculation AI 74 and the four-wheel sprung velocity calculation unit 75 constitute a sprung velocity acquisition unit that calculates the sprung velocity of each wheel. At this time, the four-wheel sprung velocity calculation unit 75 receives the sprung velocity at the IMU sensor position output from the sprung velocity calculation AI 74 and the roll rate and pitch rate at the IMU sensor position acquired by the receiving unit 16 from the CAN 10, and calculates the sprung velocity of each wheel of the vehicle. Specifically, the four-wheel sprung velocity calculation unit 75 calculates the sprung velocities for all four wheels based on the sprung velocity, roll rate, and pitch rate at the sensor position, the IMU sensor position, and vehicle specification information such as the shape, size, weight, wheelbase, and position of each wheel 3 of the vehicle body 2. In other words, if the sprung body 2 is considered to be a rigid body, the absolute vertical sprung velocity (sprung velocity) of each wheel can be geometrically calculated based on the sprung velocity at the sensor position, the roll rate, and the pitch rate.

[0100] The front wheel sprung acceleration calculation unit 76 is a sprung acceleration acquisition unit that calculates the sprung acceleration of each wheel of the vehicle 1 based on the values ​​input to the receiving unit 16 and the specifications of the vehicle 1. The front wheel sprung acceleration calculation unit 76 calculates the vertical sprung acceleration of each wheel based on the data that the receiving unit 16 acquires from the CAN 10, the sensor position of the IMU, and vehicle specification information such as the shape, size, weight, wheelbase, and position of each wheel 3 of the vehicle body 2. Specifically, the front wheel sprung acceleration calculation unit 76 calculates the sprung acceleration (vertical acceleration), roll rate, and pitch rate at the sensor position of the IMU from the data acquired from CAN 10, and also calculates the vertical sprung acceleration of the front wheels based on the sprung acceleration, roll rate, and pitch rate at the sensor position and the relationship between the sensor position and the position of the front wheels (tire positions). The front wheel sprung acceleration calculation unit 76 constitutes an IMU signal conversion unit that converts the signal output from the IMU (IMU signal) into the sprung acceleration at the tire position.

[0101] The front wheel unsprung velocity calculation AI77 is an artificial intelligence model constructed by machine learning. The front wheel unsprung velocity calculation AI77 receives the front wheel sprung acceleration from the front wheel sprung acceleration calculation unit 76 and the front wheel command current value (i), and calculates the front wheel unsprung velocity from these signals. Specifically, the front wheel unsprung velocity calculation AI77 receives the front wheel sprung acceleration and the front wheel command current value (i), and estimates the front wheel unsprung velocity using a neural network that has been trained to estimate the front wheel unsprung velocity. The front wheel unsprung velocity calculation AI77 reads the weight parameters stored in the weight parameter storage unit 14 via the data reading unit 15. The front wheel unsprung velocity calculation AI77 estimates the front wheel unsprung velocity based on the front wheel sprung acceleration, the command current value (i), and the weight parameters.

[0102] The front wheel unsprung velocity calculation AI 77 and the four-wheel stroke velocity calculation unit 78 constitute a stroke velocity acquisition unit that calculates the stroke velocities of the suspension devices 5 for four wheels using a mathematical model of the vehicle 1 that receives as input the sprung acceleration of each wheel and the control command value (command current value) for the variable damper 7 of each wheel. The four-wheel stroke velocity calculation unit 78 receives the sprung velocities for four wheels from the four-wheel sprung velocity calculation unit 75. The four-wheel stroke velocity calculation unit 78 also receives the unsprung velocity of the front wheels from the front wheel unsprung velocity calculation AI 77. At this time, the four-wheel stroke velocity calculation unit 78 calculates the unsprung velocity of the rear wheels from the unsprung velocity of the front wheels, taking into account a time delay from the unsprung velocity of the front wheels. Specifically, the unsprung velocity of the rear wheels is calculated from the unsprung velocity of the front wheels, taking into account a delay time or distance corresponding to the wheelbase. The four-wheel stroke speed calculation unit 78 calculates the stroke speeds of the suspension devices 5 for four wheels by subtracting the unsprung speed from the sprung speed of each wheel thus obtained.

[0103] The stroke speeds for four wheels calculated by the four-wheel stroke speed calculation unit 78 and the sprung speeds for four wheels calculated by the four-wheel sprung speed calculation unit 75 are input to the ride comfort control unit 49. The ride comfort control unit 49 calculates a control command value (command current value) for controlling the damping force of the variable damper 7 of the suspension device 5, based on the instantaneous values ​​input from the four-wheel stroke speed calculation unit 78 and the four-wheel sprung speed calculation unit 75.

[0104] Thus, the fifth embodiment can achieve substantially the same effects as the third embodiment. In the fifth embodiment, instead of calculating the sprung velocities for all four wheels using AI, the sprung velocities are calculated using AI from the sprung acceleration (vertical acceleration) at one location, and the sprung velocities for all four wheels are calculated using this value, the roll rate, and the pitch rate. This reduces the calculation load for the sprung velocities by one-fourth. In this case, because the IMU directly detects the velocity components of the roll rate and pitch rate, there is no need to perform processing that takes integral error into account. Furthermore, in the fifth embodiment, instead of calculating the stroke velocities for all four wheels, the unsprung velocities of the front wheels are calculated, and the unsprung velocities of the rear wheels are calculated taking into account the time delay from the front wheels. This reduces the calculation load for the stroke speed by half. As a result, the configuration of the controller 71 can be simplified, and manufacturing costs can be reduced.

[0105] 16 and 17 show a sixth embodiment. The sixth embodiment is characterized in that the controller calculates the stroke speed of the suspension device based on information from a relative displacement sensor that detects the relative displacement between the vehicle body and the wheels. In the sixth embodiment, the same components as those in the first embodiment are designated by the same reference numerals, and their description will be omitted.

[0106] FIG. 16 shows a vehicle 1 to which a controller 91 according to a sixth embodiment is applied. The controller 91 according to the sixth embodiment is a control device that controls the variable damper 7 (force generating device). The controller 91 is configured, for example, by a microcomputer as a control device that controls the damping characteristics of the variable damper 7. The controller 91 according to the sixth embodiment is configured similarly to the controller 11 according to the first embodiment. The controller 91 acquires various data from the IMU 9 and the CAN 10. In addition, the controller 91 acquires various data from a relative displacement sensor 90. The output side of the controller 91 is connected to the variable damping force actuator 8 of the variable damper 7.

[0107] The relative displacement sensor 90 is, for example, a stroke sensor or a vehicle height sensor, and is attached to the suspension device 5, the variable damper 7, etc. The relative displacement sensor 90 detects the relative displacement between the vehicle body 2 and the wheels 3. The relative displacement sensor 90 is provided on either the left front wheel or the right front wheel. Note that the relative displacement sensor 90 is not limited to being provided on the left front wheel or the right front wheel, but may also be provided on either the left rear wheel or the right rear wheel.

[0108] The controller 91 has a storage unit 12 including a ROM, a RAM, a non-volatile memory, etc. The storage unit 12 of the controller 91 stores various programs, information (sensor positions, vehicle specification information), data, etc. for controlling the variable damper 7. The controller 91 estimates the sprung velocity and stroke velocity as vehicle state quantities based on data acquired from the IMU 9 and the relative displacement sensor 90. The controller 91 determines the force to be generated by the variable damper 7 (force generating device) of the suspension device 5 based on the estimated vehicle state quantities, and outputs a control signal (command current) to the variable damping force actuator 8 of the suspension device 5.

[0109] 16 and 17 , the controller 91 includes a damper control unit 92, a weight parameter storage unit 14, and a data reading unit 15. The damper control unit 92 includes a receiving unit 16, a vehicle state quantity calculation unit 93, and a ride comfort control unit 22.

[0110] The receiving unit 16 acquires data on sprung acceleration, roll rate, and pitch rate from the IMU 9. The receiving unit 16 acquires data related to vehicle behavior (hereinafter referred to as behavior information) via the CAN 10. In addition, the receiving unit 16 acquires data on the relative displacement between the vehicle body 2 and the wheels 3 from the relative displacement sensor 90. Note that the controller 91 does not need to acquire data directly from the IMU 9 and the relative displacement sensor 90, and may acquire the data via the CAN 10.

[0111] The vehicle state quantity calculation unit 93 calculates (estimates) stroke speed and sprung speed as vehicle state quantities based on, for example, data acquired from the IMU 9 and the relative displacement sensor 90. The vehicle state quantity calculation unit 93 includes a sprung acceleration calculation unit 18, a stroke speed calculation AI 94, and a sprung speed calculation AI 20. Based on data acquired from the IMU 9 and the relative displacement sensor 90, the vehicle state quantity calculation unit 93 estimates data related to suspension control required by the ride comfort control unit 22, specifically, instantaneous values ​​related to suspension control (sprung speed, stroke speed).

[0112] The stroke speed calculation AI 94 is a stroke speed acquisition unit that calculates the stroke speed of the suspension device 5 of each wheel using a mathematical model of the vehicle 1, which inputs values ​​indicating the sprung acceleration of each wheel, the relative displacement of each wheel, and the force (damping force) generated by the variable damper 7. In this case, the stroke speed calculation AI 94 is an artificial intelligence model constructed by machine learning. The stroke speed calculation AI 94 is configured by a long short term memory (LSTM), which is a type of recurrent neural network (RNN).

[0113] An LSTM has a memory cell (not shown). An LSTM can perform machine learning while retaining some of the input information. An LSTM has a forget gate, an input gate, and an output gate (none of which are shown). The forget gate determines how much data stored in the memory cell to forget based on the input information. The input gate determines which information from the input information should be newly stored in the memory cell. The output gate determines the information to be finally output. An LSTM learns using three mechanisms: a forget gate, an input gate, and an output gate.

[0114] The stroke speed calculation AI 94 receives the sprung acceleration of each wheel output from the sprung acceleration calculation unit 18, one relative displacement output from the receiver 16, and the command current value (i) output from the ride comfort control unit 22. The stroke speed calculation AI 94 calculates the stroke speed of the suspension device 5 of each wheel based on the sprung acceleration, the relative displacement, and the command current value (i). Specifically, the stroke speed calculation AI 94 receives the sprung acceleration and relative displacement of each wheel and the command current value (i) for the variable damper 7, and estimates the stroke speed of each wheel using a neural network that has been trained to estimate the stroke speed of each wheel. The stroke speed calculation AI 94 reads out weight parameters stored in the weight parameter storage unit 14 via the data reading unit 15. The stroke speed calculation AI 94 estimates the stroke speed based on the sprung acceleration, the relative displacement, the command current value (i), and the weight parameters.

[0115] The stroke speed output by the stroke speed calculation AI 94 and the sprung speed output by the sprung speed calculation AI 20 are input to the ride comfort control unit 22. The ride comfort control unit 22 calculates a control command value (command current value) for controlling the damping force of the variable damper 7 of the suspension device 5 based on the instantaneous values ​​of the stroke speed and the sprung speed.

[0116] Thus, the sixth embodiment can achieve substantially the same effects as the first embodiment. In the first embodiment, the stroke speed (relative speed) was estimated without any information on the unsprung mass, and therefore the accuracy of the stroke speed estimation was not high. In contrast, in the sixth embodiment, the controller 91 (controller) determines the stroke speed of the suspension device 5 (shock absorber) based on information from a relative displacement sensor 90 that detects the relative displacement between the body 2 and the wheels 3 of the vehicle 1. Specifically, the controller 91 according to the sixth embodiment estimates the stroke speed using data from the relative displacement sensor 90. This makes it possible to estimate the stroke speed (relative speed) taking into account the "unsprung mass vibration component," thereby improving the accuracy of the stroke speed estimation.

[0117] As a result, the number of elements in the hidden layer of the neural network and the number of layers in the LSTM required for the target accuracy can be reduced. This allows for a reduction in processing time and the required ROM and RAM capacity. Furthermore, since the size of the neural network can be reduced, the learning time of the neural network can be shortened. In addition, since the relative displacement vibration is learned taking into account the unsprung vibration, less learning data is required.

[0118] The stroke speed calculation AI 94 (stroke speed acquisition unit) according to the sixth embodiment uses an LSTM to determine the stroke speed of the suspension device 5 (shock absorber). Generally, relative displacement (stroke) is differentiated to convert it into stroke speed. Because noise components are amplified during differentiation, an LPF (low-pass filter) is typically used to remove the noise components. However, the filter causes phase lag and attenuation. In contrast, the stroke speed calculation AI 94 according to the sixth embodiment uses an LSTM to handle relative displacement information without differentiating it. This solves the above-mentioned problems.

[0119] The relationship between the current value and the relative velocity is usually learned to correct the change in relative velocity due to the variable damping characteristics of the variable damper 7. In contrast, the stroke speed calculation AI 94 according to the sixth embodiment can correct the change in relative velocity due to the variable damping characteristics by using relative displacement information instead of the relationship between the current value and the relative velocity.

[0120] In the sixth embodiment, one relative displacement sensor 90 is provided on either the left or right front wheel of the vehicle 1, and the stroke speed calculation AI 94 (stroke speed acquisition unit) calculates the stroke speed of the suspension device 5 (shock absorber) for each wheel of the vehicle 1 from one piece of relative displacement information. Specifically, one relative displacement sensor 90 is provided on the front wheel side of the vehicle 1, and the stroke speed calculation AI 94 calculates the stroke speed for each wheel 3 (the remaining three locations) of the vehicle 1 from one piece of relative displacement information. In this case, the RNN (LSTM) constituting the stroke speed calculation AI 94 learns the correlation, including temporal changes, between the "data from the relative displacement sensor 90 for one wheel" and the "relative speed of the other wheels." This allows for improved stroke speed estimation accuracy even for three wheels not equipped with a relative displacement sensor 90. In this way, a signal from at least one relative displacement sensor 90 is sufficient to improve the stroke speed estimation accuracy.

[0121] 16 and 17 show a seventh embodiment. The seventh embodiment is characterized in that in addition to one relative displacement sensor being provided on one of the front wheels of the vehicle, another is provided on a rear wheel diagonally opposite the relative displacement sensor on the front wheel, and the stroke speed calculation AI calculates the stroke speed of the remaining variable damper of the vehicle from the two pieces of relative displacement information. In the seventh embodiment, the same components as those in the first and sixth embodiments are designated by the same reference numerals, and their description will be omitted.

[0122] 16 shows a vehicle 1 to which a controller 101 according to the seventh embodiment is applied. The controller 101 according to the seventh embodiment is a control device that controls the variable damper 7 (force generating device). The controller 101 is configured, for example, by a microcomputer as a control device that controls the damping characteristics of the variable damper 7. The controller 101 according to the seventh embodiment is configured similarly to the controller 11 according to the first embodiment. The controller 101 acquires various data from the IMU 9, the CAN 10, and the relative displacement sensors 90 and 100. The output side of the controller 101 is connected to the damping force variable actuator 8 of the variable damper 7.

[0123] The relative displacement sensors 90, 100 are stroke sensors, vehicle height sensors, etc., and detect the relative displacement between the vehicle body 2 and the wheels 3. The relative displacement sensor 90 is provided, for example, on the left front wheel. The relative displacement sensor 100 is provided on the right rear wheel so as to be diagonally opposite the relative displacement sensor 90 on the left front wheel. Alternatively, the relative displacement sensor 90 may be provided on the right front wheel and the relative displacement sensor 100 may be provided on the left rear wheel.

[0124] The controller 101 has a storage unit 12 including a ROM, a RAM, a non-volatile memory, etc. The storage unit 12 of the controller 101 stores various programs, information (sensor positions, vehicle specification information), data, etc. for controlling the variable damper 7. The controller 101 estimates the sprung velocity and stroke velocity as vehicle state quantities based on data acquired from the IMU 9 and the relative displacement sensors 90 and 100. The controller 101 calculates the force to be generated by the variable damper 7 (force generating device) of the suspension unit 5 based on the estimated vehicle state quantities, and outputs a control signal (command current) to the variable damping force actuator 8 of the suspension unit 5.

[0125] 16 and 17 , the controller 101 includes a damper control unit 102, a weight parameter storage unit 14, and a data reading unit 15. The damper control unit 102 includes a receiving unit 16, a vehicle state quantity calculation unit 103, and a ride comfort control unit 22.

[0126] The receiving unit 16 acquires data on sprung acceleration, roll rate, and pitch rate from the IMU 9. The receiving unit 16 acquires data related to vehicle behavior (hereinafter referred to as behavior information) via the CAN 10. In addition, the receiving unit 16 acquires data on two relative displacements between the vehicle body 2 and the wheels 3 from the relative displacement sensors 90 and 100. Note that the controller 101 does not need to acquire data directly from the IMU 9 and the relative displacement sensors 90 and 100, and may acquire the data via the CAN 10.

[0127] The vehicle state quantity calculation unit 103 calculates (estimates) stroke speed and sprung speed as vehicle state quantities based on data acquired from, for example, the IMU 9 and the relative displacement sensors 90 and 100. The vehicle state quantity calculation unit 103 includes a sprung acceleration calculation unit 18, a stroke speed calculation AI 104, and a sprung speed calculation AI 20. Based on data acquired from the IMU 9 and the relative displacement sensors 90 and 100, the vehicle state quantity calculation unit 103 estimates data related to suspension control required by the ride comfort control unit 22, specifically, instantaneous values ​​related to suspension control (sprung speed, stroke speed).

[0128] The stroke speed calculation AI 104 is a stroke speed acquisition unit that calculates the stroke speed of the suspension device 5 of each wheel using a mathematical model of the vehicle 1, which inputs values ​​indicating the sprung acceleration of each wheel, the relative displacement of each wheel, and the force (damping force) generated by the variable damper 7. In this case, the stroke speed calculation AI 104 is an artificial intelligence model constructed by machine learning. The stroke speed calculation AI 104 is configured by an LSTM, which is a type of recurrent neural network.

[0129] The stroke speed calculation AI 104 receives the sprung acceleration of each wheel output from the sprung acceleration calculation unit 18, the two relative displacements output from the receiver 16, and the command current value (i) output from the ride comfort control unit 22. The stroke speed calculation AI 104 calculates the stroke speed of the suspension device 5 of each wheel based on the sprung acceleration, the relative displacement, and the command current value (i). Specifically, the stroke speed calculation AI 104 receives the sprung acceleration and relative displacement of each wheel and the command current value (i) for the variable damper 7, and estimates the stroke speed of each wheel using a neural network trained to estimate the stroke speed of each wheel. The stroke speed calculation AI 104 reads the weight parameters stored in the weight parameter storage unit 14 via the data reading unit 15. The stroke speed calculation AI 104 estimates the stroke speed based on the sprung acceleration, the relative displacement, the command current value (i), and the weight parameters.

[0130] The stroke speed output by the stroke speed calculation AI 104 and the sprung speed output by the sprung speed calculation AI 20 are input to the ride comfort control unit 22. The ride comfort control unit 22 calculates a control command value (command current value) for controlling the damping force of the variable damper 7 of the suspension device 5 based on the instantaneous values ​​of the stroke speed and the sprung speed.

[0131] Thus, the seventh embodiment can achieve substantially the same effects as the first embodiment. In the seventh embodiment, the relative displacement sensor 100 is provided on the rear wheel diagonally opposite the relative displacement sensor 90 on the front wheel, and the stroke speed calculation AI 104 (stroke speed acquisition unit) calculates the stroke speeds of the remaining suspension devices 5 (shock absorbers) of the vehicle 1 from the two pieces of relative displacement information. Specifically, one relative displacement sensor 90 is provided on the left front wheel (or right front wheel), and one relative displacement sensor 100 is provided on the right rear wheel (or left rear wheel). The stroke speed calculation AI 104 calculates the stroke speeds of the remaining wheels 3 (the remaining two locations) of the vehicle 1 from the two pieces of relative displacement information. In this case, the RNN (LSTM) constituting the stroke speed calculation AI 104 learns the correlation, including temporal changes, between the "relative displacement sensor data for two wheels" and the "relative speeds of the other wheels." This improves the accuracy of stroke speed estimation even for two wheels not provided with relative displacement sensors 90, 100. By using the signals from the two relative displacement sensors 90, 100, the stroke speed can be estimated with higher estimation accuracy than when using only one relative displacement sensor 90. In particular, by arranging the relative displacement sensors 90, 100 at positions where the front and rear are diagonally opposite each other, the signal from the IMU 9 can be applied as long as it contains one of the values ​​of vertical acceleration, roll rate, and pitch rate.

[0132] In the seventh embodiment, the relative displacement sensor 100 is mounted on the rear wheel, thereby improving the accuracy of estimating the stroke speed of the rear wheel. Furthermore, the learning speed of the stroke speed calculation AI 104 is improved compared to when the relative displacement sensors 90 and 100 are omitted. This shortens the learning time required to reach a certain level of accuracy. This facilitates application to real-time learning. It may be possible to reduce the number of elements in the hidden layer of the neural network and the number of layers in the LSTM, thereby reducing processing time and the required ROM and RAM capacities, and enabling the use of inexpensive microcomputers.

[0133] In the seventh embodiment, the relative displacement sensor 100 is provided on the rear wheel diagonally opposite the relative displacement sensor 90 on the front wheel, and the stroke speed calculation AI 104 (stroke speed acquisition unit) calculates the stroke speed of the remaining suspension device 5 (shock absorber) of the vehicle 1 from the two pieces of relative displacement information. However, the present invention is not limited to this. For example, one relative displacement sensor may be provided on either the left or right front wheel, and another may be provided on either the other front wheel or the rear wheel, and the stroke speed acquisition unit may calculate the stroke speed of the remaining shock absorber of the vehicle from the two pieces of relative displacement information. That is, the relative displacement sensors may be provided on the left front wheel and the right front wheel. Alternatively, the relative displacement sensors may be provided on the left front wheel and the left rear wheel, or the right front wheel and the right rear wheel. In this case, the stroke speed acquisition unit calculates the stroke speed of each of the remaining wheels (the remaining two locations) of the vehicle from the two pieces of relative displacement information. Even with this configuration, the same effects as those of the seventh embodiment can be obtained.

[0134] Relative displacement sensors may be attached to three or more wheels. Installing relative displacement sensors on three or more wheels also improves estimation accuracy. At least three sensors are required to calculate the three-dimensional rotational motion of the vehicle. When considering combination with an IMU, providing relative displacement sensors at two diagonal positions of the four wheels, as in the seventh embodiment, can reduce the number of relative displacement sensors and reduce costs while ensuring the estimation accuracy of the stroke speed.

[0135] In the sixth and seventh embodiments, the stroke speed calculation AI 94, 104 calculates the stroke speed of the suspension device 5 of each wheel based on the sprung acceleration, relative displacement, and damping force, but the present invention is not limited to this. For example, the stroke speed calculation AI may calculate the stroke speed of the suspension device 5 of each wheel based on the sprung velocity, relative displacement, and damping force. The stroke speed calculation AI 94, 104 is not limited to an LSTM, and may be configured using various types of neural networks.

[0136] 18 and 19 show an eighth embodiment. The eighth embodiment is characterized in that the controller includes a four-wheel stroke calculation AI (stroke acquisition unit) that calculates the stroke (relative displacement) of the suspension device (shock absorber) of each wheel using a mathematical model of the vehicle that inputs values ​​indicating the sprung acceleration of each wheel and the force generated by the variable damper, and a four-wheel stroke speed calculation unit (stroke speed acquisition unit) that calculates the stroke of the suspension device of each wheel and calculates the stroke speed of the suspension device of each wheel. In the eighth embodiment, the same components as those in the first and fifth embodiments are designated by the same reference numerals, and their description will be omitted.

[0137] 18 shows a vehicle 1 to which a controller 111 according to the eighth embodiment is applied. The controller 111 according to the eighth embodiment is a control device that controls the variable damper 7 (force generating device). The controller 111 is configured, for example, by a microcomputer as a control device that controls the damping characteristics of the variable damper 7. The controller 111 according to the eighth embodiment is configured similarly to the controller 11 according to the first embodiment. The controller 111 acquires various data from the IMU 9 and the CAN 10. The output side of the controller 111 is connected to the damping force variable actuator 8 of the variable damper 7.

[0138] The controller 111 has a storage unit 12 including a ROM, a RAM, a non-volatile memory, etc. The storage unit 12 of the controller 111 stores various programs, information (sensor positions, vehicle specification information), data, etc. for controlling the variable damper 7. The controller 111 estimates the sprung velocity and stroke velocity as vehicle state quantities based on data acquired from the IMU 9 and the CAN 10. The controller 111 determines the force to be generated by the variable damper 7 (force generating device) of the suspension device 5 based on the estimated vehicle state quantities, and outputs the control signal (command current) to the variable damping force actuator 8 of the suspension device 5.

[0139] 18 and 19 , the controller 111 includes a damper control unit 112, a weight parameter storage unit 14, and a data reading unit 15. The damper control unit 112 includes a receiving unit 16, a vehicle state quantity calculation unit 113, and a control quantity determination unit 117.

[0140] The vehicle state quantity calculation unit 113 calculates (estimates) stroke speed and sprung speed as vehicle state quantities based on data acquired from, for example, the IMU 9 and the CAN 10. The vehicle state quantity calculation unit 113 includes a sprung speed calculation unit 74, a four-wheel sprung speed calculation unit 75, a four-wheel sprung acceleration calculation unit 114, a four-wheel stroke calculation AI 115, and a four-wheel stroke speed calculation unit 116. The vehicle state quantity calculation unit 113 estimates data related to suspension control required by the ride comfort control unit 49, specifically, instantaneous values ​​related to suspension control (sprung speed, stroke speed), based on data acquired from the IMU 9 and the CAN 10.

[0141] The sprung velocity calculation AI 74 and the four-wheel sprung velocity calculation unit 75 constitute a sprung velocity acquisition unit that calculates the sprung velocity of each wheel. At this time, the four-wheel sprung velocity calculation unit 75 receives the sprung velocity at the sensor position of the IMU 9 output from the sprung velocity calculation AI 74 and the roll rate and pitch rate at the sensor position of the IMU 9 acquired by the receiving unit 16 from the CAN 10, and calculates the sprung velocity of each wheel of the vehicle. Specifically, the four-wheel sprung velocity calculation unit 75 calculates the sprung velocities for all four wheels based on the sprung velocity, roll rate, and pitch rate at the sensor position, the sensor position of the IMU 9, and vehicle specification information such as the shape, size, weight, wheelbase, and position of each wheel 3 of the vehicle body 2.

[0142] The four-wheel sprung acceleration calculation unit 114 is a sprung acceleration acquisition unit that calculates the sprung acceleration of each wheel of the vehicle 1 based on the values ​​input to the receiving unit 16 and the specifications of the vehicle 1. The four-wheel sprung acceleration calculation unit 114 calculates the vertical sprung acceleration of each wheel based on the data acquired by the receiving unit 16 from the IMU 9 and CAN 10, the sensor position of the IMU 9, and vehicle specification information such as the shape, size, weight, wheelbase, and position of each wheel 3 of the vehicle body 2. Specifically, the four-wheel sprung acceleration calculation unit 114 calculates the sprung acceleration (vertical acceleration), roll rate, and pitch rate at the sensor position of the IMU 9 from the data acquired from the IMU 9 and CAN 10, and calculates the vertical sprung acceleration for all four wheels based on the sprung acceleration, roll rate, and pitch rate at the sensor position and the relationship between the sensor position and the positions of the four wheels (tire positions). The four-wheel sprung acceleration calculation unit 114 constitutes an IMU signal conversion unit that converts the signal (IMU signal) output from the IMU 9 into the sprung acceleration at the tire position.

[0143] The four-wheel stroke calculation AI 115 receives the sprung accelerations of the four wheels output from the four-wheel sprung acceleration calculation unit 114 and the command current values ​​(i) for the four wheels, and calculates the strokes of the four wheels of the vehicle from these signals. The four-wheel stroke calculation AI 115 is a stroke acquisition unit that obtains the stroke (relative displacement) of the suspension device 5 of each wheel using a mathematical model of the vehicle 1 that receives as input the sprung acceleration of each wheel and values ​​indicating the force (damping force) generated by the variable damper 7. In this case, the four-wheel stroke calculation AI 115 is an artificial intelligence model constructed by machine learning. The four-wheel stroke calculation AI 115 receives as input the sprung accelerations of each wheel and the command current values ​​(i) for each wheel, and estimates the stroke of the suspension device 5 of each wheel using a neural network that has been trained to estimate the stroke of each wheel.

[0144] The four-wheel stroke speed calculation unit 116 is a stroke speed acquisition unit that performs calculations on the stroke of the suspension unit 5 of each wheel and obtains the stroke speed of the suspension unit 5 of each wheel. The four-wheel stroke speed calculation unit 116 is configured with, for example, a differential filter. The four-wheel stroke speed calculation unit 116 differentiates the stroke of the suspension unit 5 of each wheel output from the four-wheel stroke calculation AI 115 to calculate the stroke speed of the suspension unit 5 of each wheel.

[0145] The control amount determination unit 117 determines the control amount for controlling the variable damper 7 based on the sprung speed and the stroke speed. The control amount determination unit 117 includes a ride comfort control unit 49, a full stroke suppression control unit 118, and a control command arbitration unit 119.

[0146] The ride comfort control unit 49 determines the control amount for controlling the variable damper 7 from the sprung speed and the stroke speed. The ride comfort control unit 49 calculates a control command value (command current value) for controlling the damping force of the variable damper 7 of the suspension device 5 based on the instantaneous values ​​input from the four-wheel sprung speed calculation unit 75 and the four-wheel stroke speed calculation unit 116.

[0147] The full stroke suppression control unit 118 calculates a damping force control signal for performing full stroke suppression control (i.e., full extension suppression control and full compression suppression control) as a command current according to the damping characteristics based on the stroke and stroke speed of each wheel. The full stroke suppression control unit 118 determines whether the variable damper 7 of each wheel is approaching a full extension state or a full compression state based on the stroke and stroke speed of each wheel. When the variable damper 7 approaches a full extension state, the full stroke suppression control unit 118 calculates a control command value (command current value) for suppressing full extension. When the variable damper 7 approaches a full compression state, the full stroke suppression control unit 118 calculates a control command value (command current value) for suppressing full compression.

[0148] The control command arbitration unit 119 outputs a control command value (command current value) for varying (adjusting) the damping force of the variable damper 7 in accordance with the control command value from the ride comfort control unit 49 and the control command value from the full stroke suppression control unit 118. The control command arbitration unit 119 arbitrates between the control command value from the ride comfort control unit 49 and the control command value from the full stroke suppression control unit 118 and outputs a final control command value. The control command arbitration unit 119 compares the control command value from the ride comfort control unit 49 with the control command value from the full stroke suppression control unit 118, selects the larger control command value, and outputs it as the final control command value (command current value). Specifically, the control command arbitration unit 119 selects the hard value from the control command value from the ride comfort control unit 49 and the control command value from the full stroke suppression control unit 118. At this time, the control command arbitration unit 119 performs the same process for the control command values ​​for each of the four wheels. A command current is supplied as a control signal based on the control command value to the damping force variable actuator 8 of the variable damper 7. In this way, the damping force of the variable damper 7 is controlled by the controller 111.

[0149] Thus, the eighth embodiment can also achieve substantially the same effects as the first embodiment. The controller 111 (control device) according to the eighth embodiment includes a receiver 16 (motion state acquirer) that acquires sprung acceleration (vertical acceleration), which is acceleration in the vertical direction, a roll rate, and a pitch rate from an IMU 9 (multipurpose sensor) that the vehicle 1 has, a four-wheel sprung acceleration calculator 114 (sprung acceleration acquirer) that calculates the sprung acceleration of each wheel of the vehicle 1 based on the values ​​input to the receiver 16 and the specifications of the vehicle 1, and a four-wheel sprung acceleration calculator 114 that calculates the stroke (relative displacement) of the suspension device 5 of each wheel using a mathematical model (AI) of the vehicle 1 that receives as input the sprung acceleration of each wheel and values ​​indicating the force generated by the variable damper 7. a four-wheel stroke speed calculation unit 116 (stroke speed acquisition unit) that performs arithmetic processing (differential filter) of the stroke of the suspension device 5 of each wheel to determine the stroke speed of the suspension device 5 of each wheel; a sprung speed calculation AI 74 and a four-wheel sprung speed calculation unit 75 (sprung speed acquisition unit) that use a mathematical model (AI) of the vehicle 1 using the values ​​input to the receiving unit 16 to determine the sprung speed of the vehicle 1; and a control amount determination unit 117 that determines the control amount for controlling the variable damper 7 from the stroke speed and sprung speed.

[0150] That is, the eighth embodiment is a control method for controlling a variable damper 7 (force generating device) of a suspension device 5 (buffer device) provided on a vehicle, and includes: a motion state acquisition step of acquiring sprung acceleration (vertical acceleration), which is acceleration in the vertical direction, a roll rate, and a pitch rate from a device (IMU 9) other than the suspension device 5 provided on the vehicle 1; a sprung acceleration acquisition step of determining the sprung acceleration of each wheel of the vehicle 1 based on the values ​​acquired in the motion state acquisition step and the specifications of the vehicle 1; a stroke acquisition step of determining the stroke of the suspension device 5 of each wheel using a mathematical model of the vehicle 1 that has as input the sprung acceleration of each wheel and values ​​indicating the force generated by the variable damper 7 (force generating device); a stroke speed acquisition step of calculating the stroke of the suspension device 5 of each wheel to determine the stroke speed of the suspension device 5 of each wheel; a sprung velocity acquisition step of determining the sprung velocity of the vehicle 1 using the value acquired in the motion state acquisition step and the mathematical model of the vehicle 1; and a control quantity determination step of determining a control quantity for controlling the variable damper 7 from the sprung velocity and the stroke velocity.

[0151] In the eighth embodiment, the sprung velocity calculation AI 74 has one input and one output, which reduces the calculation load of the sprung velocity calculation AI 74. Furthermore, in the first to seventh embodiments, the stroke velocity (relative velocity) is estimated from the sprung velocity, etc. In this case, it is also possible to calculate the relative displacement by integrating the relative velocity estimated by the AI. However, compared to differential processing, integral processing increases the gain of extremely low frequencies, making the calculation more difficult. In particular, to perform full-stroke suppression control, improved stroke (relative displacement) accuracy is required. Furthermore, when using five existing acceleration sensors, the relative displacement is calculated by double-integrating acceleration, resulting in poor estimation accuracy of the relative displacement.

[0152] In contrast, in the eighth embodiment, the stroke (relative displacement) is estimated by the four-wheel stroke calculation AI 115, and the estimated stroke is subjected to arithmetic processing using a differential filter to calculate the stroke speed. As a result, in the eighth embodiment, the stroke is estimated by AI, so the stroke and stroke speed can be estimated with high accuracy. As a result, the stroke and stroke speed can be input to the full stroke suppression control unit 118 with high accuracy, allowing for appropriate full stroke suppression control.

[0153] In the first embodiment, the stroke speed calculation AI19 calculates the stroke speed of the suspension device 5 of each wheel based on the sprung acceleration and the damping force, but the present invention is not limited to this. As in the modified example shown in FIG. 15 , the stroke speed calculation AI83 may calculate the stroke speed of the suspension device 5 of each wheel based on the sprung acceleration and the damping force. A damper control unit 81 according to the modified example shown in FIG. 15 includes a receiving unit 16, a vehicle state quantity calculation unit 82, and a ride comfort control unit 22. The vehicle state quantity calculation unit 82 includes a sprung acceleration calculation unit 18 and a stroke speed calculation AI83. In addition, the vehicle state quantity calculation unit 82 includes the integrator 35 according to the second embodiment. In this case, the integrator 35 may be replaced with the sprung speed calculation AI20 according to the first embodiment. The stroke speed calculation AI83 according to the modified example can also be applied to the second to fourth embodiments. In the first embodiment, the stroke speed calculation AI 19 is configured using an artificial intelligence model (mathematical model) equipped with a neural network, but the present invention is not limited to this. As long as the correlation between input and output can be learned, the stroke speed calculation AI 19 may be configured using a numerical model without a neural network. This configuration can also be applied to the sprung speed calculation AIs 20, 46, and 74, the stroke speed calculation AIs 34, 45, 64, 83, 94, and 104, the front wheel unsprung speed calculation AI 77, and the four-wheel stroke calculation AI 115 according to the first to eighth embodiments.

[0154] In the third embodiment, the control variable determiner 48 includes the ride comfort control unit 49 and the driving stability control unit 50, but the present invention is not limited to this. The control variable determiner 48 may omit the driving stability control unit 50. In other words, the control variable determiner 48 may be configured with only the ride comfort control unit 49. This configuration can also be applied to the fourth embodiment.

[0155] In the third embodiment, the controller 41 acquires data from the IMU through the CAN 10, but the present invention is not limited to this. For example, the controller may acquire data from the IMU directly from the IMU, as in the first embodiment. This configuration can also be applied to the fourth embodiment.

[0156] In the first embodiment, the controller 11 acquires vehicle information including wheel speeds via the CAN 10, but the present invention is not limited to this. For example, the controller 11 may acquire detected values ​​of various sensors directly from the various sensors. Furthermore, the controller 11 may acquire behavior information from another controller, etc. This configuration can also be applied to the second, third, fourth, sixth, seventh, and eighth embodiments.

[0157] In the first to eighth embodiments, the IMU is installed outside the controller and receives and controls data from the controller. However, the present invention is not limited to this, and the IMU may be installed inside the controller.

[0158] In the above-described embodiments, the force generating device is a variable damper 7 made of a semi-active damper. However, the present invention is not limited to this, and an active damper (either an electric actuator or a hydraulic actuator) may be used as the force generating device. In the above-described embodiments, the force generating device that generates a force that can be adjusted between the vehicle body 2 side and the wheel 3 side is configured as a variable damper 7 made of a damping force adjustable hydraulic shock absorber. However, the present invention is not limited to this, and the force generating device may be configured as an air suspension, a stabilizer (kinesus), an electromagnetic suspension, or the like, in addition to a hydraulic shock absorber.

[0159] In the above embodiments, the control device for a suspension device used in a four-wheeled automobile has been described as an example. However, the present invention is not limited to this and may be applied to, for example, two-wheeled and three-wheeled vehicles, as well as work vehicles and transport vehicles such as trucks and buses.

[0160] The present invention is not limited to the above-described embodiments and includes various modifications. For example, the above-described embodiments have been described in detail to clearly explain the present invention, and the present invention is not necessarily limited to those including all of the described configurations. Furthermore, it is possible to replace part of the configuration of one embodiment with the configuration of another embodiment, or to add the configuration of another embodiment to the configuration of one embodiment. Furthermore, it is possible to add, delete, or replace part of the configuration of each embodiment with other configurations.

[0161] 1: Vehicle, 2: Vehicle body, 3: Wheel, 5: Suspension device (buffer device), 7: Variable damper (force generating device), 8: Variable damping force actuator, 9: IMU (multipurpose sensor), 10: CAN, 11, 31, 41, 61, 71, 91, 101, 111: Controller, 13, 32, 42, 62, 72, 81, 92, 102, 112: Damper control unit, 16: Receiver (motion state acquisition unit), 18, 44: Spring acceleration calculation unit (sprung acceleration acquisition unit), 19, 34, 45, 64, 83, 94, 104: Stroke speed calculation AI (stroke speed acquisition unit), 20, 46: Spring velocity calculation AI (sprung velocity acquisition unit), 21, 47: Damping force calculation unit (damping damping force acquisition unit), 22: ride comfort control unit (control amount determination unit), 35: integrator (sprung velocity acquisition unit), 48, 117: control amount determination unit, 49: ride comfort control unit, 50: handling stability control unit, 51: control command management unit, 74: sprung velocity calculation AI, 75: 4-wheel sprung velocity calculation unit (sprung velocity acquisition unit), 76: front wheel sprung acceleration calculation unit (sprung acceleration acquisition unit), 77: front wheel unsprung velocity calculation AI, 78, 116: 4-wheel stroke velocity calculation unit (stroke velocity acquisition unit), 114: 4-wheel sprung acceleration calculation unit (sprung acceleration acquisition unit), 115: 4-wheel stroke calculation AI (stroke acquisition unit), 118: full stroke suppression control unit, 119: control command arbitration unit

Claims

1. A control device for controlling a force generating device of a shock absorber provided on a vehicle, comprising: a motion state acquisition unit that acquires vertical acceleration, which is acceleration in the vertical direction, a roll rate, and a pitch rate from a multipurpose sensor provided on the vehicle; a sprung acceleration acquisition unit that calculates the sprung acceleration of each wheel of the vehicle based on values ​​input to the motion state acquisition unit and specifications of the vehicle; a stroke speed acquisition unit that calculates the stroke speed of the shock absorber of each wheel using a mathematical model of the vehicle that inputs the sprung acceleration of each wheel and values ​​indicating the force generated by the force generating device; a sprung speed acquisition unit that calculates the sprung speed of each wheel; and a control quantity determination unit that determines a control quantity for controlling the force generating device from the sprung speed and the stroke speed.

2. The control device according to claim 1, wherein the sprung velocity acquisition unit is an artificial intelligence model constructed by machine learning.

3. A control device according to claim 1, wherein the stroke speed acquisition unit is an artificial intelligence model constructed by machine learning.

4. A control device according to claim 2, wherein the stroke speed acquisition unit is an artificial intelligence model constructed by machine learning.

5. A control device according to claim 1, further comprising a damping force acquisition unit that acquires a damping force using the control amount as an input as a value indicating the force generated by the force generating device, and feeds the damping force back into the stroke speed acquisition unit.

6. A control method for controlling a force generating device of a shock absorber provided on a vehicle, comprising: a motion state acquisition step of acquiring vertical acceleration, which is acceleration in the vertical direction, a roll rate, and a pitch rate from a device other than the shock absorber provided on the vehicle; a sprung acceleration acquisition step of determining the sprung acceleration of each wheel of the vehicle based on the values ​​acquired in the motion state acquisition step and specifications of the vehicle; a stroke speed acquisition step of determining the stroke speed of the shock absorber of each wheel using a mathematical model of the vehicle that inputs the sprung acceleration of each wheel and values ​​indicating the force generated by the force generating device; a sprung velocity acquisition step of determining the sprung velocity of each wheel; and a control quantity determination step of determining a control quantity for controlling the force generating device from the sprung velocity and the stroke velocity.

7. A control device according to claim 1, wherein the stroke speed acquisition unit is provided on the vehicle and determines the stroke speed of the shock absorber based on information from a relative displacement sensor that detects relative displacement between the body and wheels of the vehicle.

8. A control device according to claim 7, wherein the stroke speed acquisition unit obtains the stroke speed of the shock absorber using an LSTM.

9. A control device according to claim 8, wherein the relative displacement sensor is provided on one of the front wheels of the vehicle, and the stroke speed acquisition unit determines the stroke speed of the shock absorber for each wheel of the vehicle from the single piece of relative displacement information.

10. A control device according to claim 9, wherein the relative displacement sensor is provided on a rear wheel diagonally opposite the sensor on the front wheel, and the stroke speed acquisition unit determines the stroke speed of the remaining shock absorber of the vehicle from the two pieces of relative displacement information.

11. A control device according to claim 9, wherein one more relative displacement sensor is provided on either the other front wheel or the rear wheel, and the stroke speed acquisition unit determines the stroke speed of the remaining shock absorber of the vehicle from the two pieces of relative displacement information.

12. A control device for controlling a force generating device of a shock absorber provided on a vehicle, comprising: a motion state acquisition unit that acquires vertical acceleration, which is acceleration in the vertical direction, a roll rate, and a pitch rate from a multipurpose sensor provided on the vehicle; a sprung acceleration acquisition unit that determines the sprung acceleration of each wheel of the vehicle based on values ​​input to the motion state acquisition unit and specifications of the vehicle; a stroke acquisition unit that determines the stroke of the shock absorber of each wheel using a mathematical model of the vehicle that has as input the sprung acceleration of each wheel and values ​​indicating the force generated by the force generating device; a stroke speed acquisition unit that performs arithmetic processing on the stroke of the shock absorber of each wheel to determine the stroke speed of the shock absorber of each wheel; a sprung velocity acquisition unit that determines the sprung velocity of each wheel using values ​​input to the motion state acquisition unit and the mathematical model of the vehicle; and a control quantity determination unit that determines a control quantity for controlling the force generating device from the sprung velocity and the stroke speed.

13. A control method for controlling a force generating device of a shock absorber provided on a vehicle, comprising: a motion state acquisition step of acquiring vertical acceleration, which is acceleration in the vertical direction, a roll rate, and a pitch rate from a device provided on the vehicle other than the shock absorber; a sprung acceleration acquisition step of determining the sprung acceleration of each wheel of the vehicle based on the values ​​acquired in the motion state acquisition step and specifications of the vehicle; a stroke acquisition step of determining the stroke of the shock absorber of each wheel using a mathematical model of the vehicle that has as input the sprung acceleration of each wheel and values ​​indicating the force generated by the force generating device; a stroke speed acquisition step of calculating the stroke of the shock absorber of each wheel and determining the stroke speed of the shock absorber of each wheel; a sprung velocity acquisition step of determining the sprung velocity of each wheel using the values ​​acquired in the motion state acquisition step and the mathematical model of the vehicle; and a control quantity determination step of determining a control quantity for controlling the force generating device from the sprung velocity and the stroke velocity.

Images