Vehicle control apparatus

Abstract

The present invention accurately estimates a vehicle state amount. This vehicle control apparatus includes: a variable damper (relative displacement control device) that controls relative displacement between a vehicle body and a wheel; a vehicle state calculation unit that calculates the state of the vehicle; a relative speed estimation unit that estimates relative speeds of wheels of the vehicle by using a neural network that has been trained so as to estimate the relative speeds of the wheels by receiving, as input, wheel speed information of the wheels outputted from the vehicle state calculation unit; a sprung speed estimation unit that estimates sprung speeds of wheels of the vehicle by using a neural network that has been trained so as to estimate the sprung speeds of the wheels by receiving, as input, wheel speed information of the wheels outputted from the vehicle state calculation unit; and a control value computing unit (control unit) that controls the variable damper on the basis of output from the relative speed estimation unit and the sprung speed estimation unit.

Description

Vehicle control device

[0001] The present disclosure relates to a vehicle control device that controls, for example, a relative displacement control device of a vehicle.

[0002] Patent Document 1 discloses a device for estimating the state of a vehicle such as a four-wheeled automobile, which uses a neural network (NN) to calculate vehicle state quantities (e.g., the relative speed between the vehicle body and the wheels, the speed of the vehicle body in the vertical direction, etc.) from wheel speeds and the like.

[0003] International Publication No. 2022 / 168683

[0004] According to the prior art of Patent Document 1, when estimating vehicle state quantities of a four-wheeled vehicle, an independent estimation device (neural network) is used for each wheel. That is, in this prior art, in order to reduce the processing load of each neural network, the sprung velocity and relative velocity of one wheel are estimated from wheel speed information for one wheel. Therefore, although the neural network receives an input signal from the wheel to be estimated (e.g., the left front wheel), it does not receive input signals from other wheels (e.g., the right front wheel, the left rear wheel, and the right rear wheel). As a result, when vibration occurs in another wheel, the estimation device according to the prior art cannot accurately estimate the effect of this vibration on the wheel to be estimated. As a result, the neural network according to the prior art needs to be trained on road surfaces with different spatial frequencies, which are road surfaces where the vibration occurrence timing of the other wheels is different, resulting in an increase in the learning pattern and the learning labor.

[0005] An object of one embodiment of the present invention is to provide a vehicle control device that can estimate vehicle state quantities with high accuracy.

[0006] A vehicle control device according to one embodiment of the present invention comprises a relative displacement control device that is provided between the vehicle body and each of a plurality of wheels and controls the relative displacement between the vehicle body and the wheels; a vehicle state calculation unit that is mounted on the vehicle and calculates the state of the vehicle; a relative speed estimation unit that receives wheel speed information of each wheel output from the vehicle state calculation unit and estimates the relative speed of each wheel using a neural network that has been trained to estimate the relative speed of each wheel of the vehicle; a sprung speed estimation unit that receives wheel speed information of each wheel output from the vehicle state calculation unit and estimates the sprung speed of each wheel using a neural network that has been trained to estimate the sprung speed of each wheel of the vehicle; and a control unit that controls a relative displacement suppression device based on outputs from the relative speed estimation unit and the sprung speed estimation unit.

[0007] A vehicle control device according to one embodiment of the present invention includes a relative displacement control device that is provided between a vehicle body and each of a plurality of wheels and controls the relative displacement between the vehicle body and the wheels; a vehicle state calculation unit that is mounted on the vehicle and calculates the state of the vehicle; a relative speed estimation unit that estimates the relative speed of each wheel using a neural network that has been trained to estimate the relative speed of each wheel of the vehicle based on wheel speed information of three wheels excluding diagonal wheels from the wheel speed information of the wheels input from the vehicle state calculation unit; a sprung speed estimation unit that estimates the sprung speed of each wheel using a neural network that has been trained to estimate the sprung speed of each wheel of the vehicle based on wheel speed information of three wheels excluding diagonal wheels from the wheel speed information of the wheels input from the vehicle state calculation unit; and a control unit that controls a relative displacement suppression device based on outputs from the relative speed estimation unit and the sprung speed estimation unit.

[0008] A vehicle control device according to one embodiment of the present invention includes a relative displacement control device that is provided between the vehicle body and each of a plurality of wheels and controls the relative displacement between the vehicle body and the wheels; a vehicle state calculation unit that is mounted on the vehicle and calculates the state of the vehicle; an unsprung speed estimation unit that receives wheel speed information of each wheel output from the vehicle state calculation unit and estimates the unsprung speed of two of the vehicle's wheels using a neural network that has been trained to estimate the unsprung speed of the two wheels; a sprung speed estimation unit that receives wheel speed information of each wheel output from the vehicle state calculation unit and estimates the sprung speed of the three wheels using a neural network that has been trained to estimate the sprung speed of three of the vehicle's wheels; and a control unit that controls a relative displacement suppression device based on outputs from the unsprung speed estimation unit and the sprung speed estimation unit.

[0009] According to one embodiment of the present invention, the vehicle state quantity can be estimated with high accuracy.

[0010] 1 is an overall configuration diagram showing a four-wheeled vehicle to which a vehicle control device according to an embodiment of the present invention is applied. FIG. 1 is a block diagram showing a vehicle control device according to an embodiment of the present invention. FIG. 2 is a block diagram showing a state estimating unit according to a first embodiment. FIG. 3 is a block diagram showing a relative speed estimating unit in FIG. 3. FIG. 4 is a block diagram showing a sprung speed estimating unit in FIG. 3. FIG. 5 is an explanatory diagram showing an example of a neural network of the relative speed estimating unit. FIG. 6 is a characteristic line diagram showing changes over time in the sprung speed of a right front wheel and a right rear wheel. FIG. 7 is a block diagram showing a state estimating unit according to a second embodiment. FIG. 8 is a block diagram showing a relative speed estimating unit in FIG. 8. FIG. 9 is a block diagram showing a state estimating unit according to a third embodiment. FIG. 11 is a block diagram showing an unsprung speed estimating unit in FIG. 11. FIG. 12 is a block diagram showing a sprung speed estimating unit in FIG. 11.

[0011] DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS A vehicle control device according to an embodiment of the present invention will be described in detail below with reference to the accompanying drawings, taking as an example a case where the vehicle control device is applied to a four-wheeled vehicle.

[0012] 1 and 2 show a vehicle control device 1. The vehicle control device 1 is composed of a suspension device 5 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 are composed of tires 4, which act as springs that absorb small irregularities in the road surface. The vehicle body 2 constitutes the sprung mass, and the wheels 3 constitute the unsprung mass.

[0013] The suspension device 5 is provided between the vehicle body 2 and the vehicle wheel 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 wheel 3.

[0014] 2 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.

[0015] 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 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.

[0016] 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.

[0017] The controller 11 constitutes a control device. The controller 11 is configured by, for example, a microcomputer, and serves as a control device for controlling the damping characteristics of the variable damper 7. The controller 11 is connected to, for example, a CAN 10 (Controller Area Network), which is a line network required for data communication. The controller 11 acquires data related to the vehicle behavior (hereinafter referred to as behavior information) through 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. Therefore, this behavior information is input to the controller 11. The output side of the controller 11 is connected to the damping force variable actuator 8 of the variable damper 7.

[0018] The controller 11 also has a storage unit 12 made up of ROM, RAM, non-volatile memory, etc. The storage unit 12 of the controller 11 stores various programs, information (vehicle information), data, etc. for controlling the variable damper 7. The controller 11 estimates the sprung velocity and relative velocity as vehicle state quantities based on the behavior information. In this case, the relative velocity is the relative velocity between the sprung and unsprung parts, and is the piston velocity of the variable damper 7. The controller 11 calculates the force to be generated by the variable damper 7 (force generating mechanism) 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.

[0019] 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 vehicle state calculation unit 16, a state estimation unit 17, and a control value calculation unit 20. The weight parameter storage unit 14 may be a part of the storage unit 12 or may be separate from the storage unit 12.

[0020] The vehicle state calculation unit 16 is mounted on the vehicle and calculates the state of the vehicle. Specifically, the vehicle state calculation unit 16 acquires data related to the vehicle state during driving from communication data transmitted within the vehicle via the CAN 10, and calculates behavior information (vehicle state) such as wheel speed, longitudinal acceleration (longitudinal G), lateral acceleration (lateral G), steering angle, and yaw rate based on the acquired data. Note that the behavior information calculated by the vehicle state calculation unit 16 from the CAN signal includes wheel speed, longitudinal acceleration (longitudinal G), lateral acceleration (lateral G), steering angle, and yaw rate, but the present invention is not limited to this. The behavior information may be information necessary for the state estimation unit 17 to estimate the sprung velocity or relative velocity, and unnecessary information may be omitted. Furthermore, the behavior information is not limited to the above-described information, and various information may be added, for example, to improve the estimation accuracy of the sprung velocity, etc.

[0021] Various types of behavior information calculated based on the CAN signal are input to the state estimation unit 17 as input signals from the vehicle state calculation unit 16. At this time, the behavior information input to the state estimation unit 17 includes wheel speeds for all four wheels.

[0022] The state estimation unit 17 includes a relative speed estimation unit 18 that estimates the relative speed of each wheel, and a sprung speed estimation unit 19 that estimates the relative speed of each wheel. The relative speed estimation unit 18 receives the wheel speed information of each wheel output from the vehicle state calculation unit 16, and estimates the relative speed of each wheel using a neural network that has been trained to estimate the relative speed of each wheel of the vehicle. The sprung speed estimation unit 19 receives the wheel speed information of each wheel output from the vehicle state calculation unit 16, and estimates the sprung speed of each wheel using a neural network that has been trained to estimate the sprung speed of each wheel of the vehicle.

[0023] The weight parameters of the trained neural network are stored in the weight parameter storage unit 14. Therefore, the neural network of the relative speed estimation unit 18 and the neural network of the sprung speed estimation unit 19 are configured using the weight parameters stored in the weight parameter storage unit 14. The neural network of the relative speed estimation unit 18 outputs relative speeds for four wheels in response to an input signal from the vehicle state calculation unit 16. The neural network of the sprung speed estimation unit 19 outputs sprung speeds for four wheels in response to an input signal from the vehicle state calculation unit 16.

[0024] 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.

[0025] The state estimation unit 17 (relative velocity estimation unit 18, sprung velocity estimation unit 19) reads out the weight parameters stored in the weight parameter storage unit 14 via the data reading unit 15. Based on the behavior information and weight parameters calculated by the vehicle state calculation unit 16, the state estimation unit 17 estimates data related to suspension control required by the control value calculation unit 20, specifically, instantaneous values ​​related to suspension control (sprung velocity, relative velocity).

[0026] The control value calculation unit 20 is a control unit that controls the variable damper 7 based on the output from the state estimation unit 17. The control value calculation unit 20 calculates a suspension control value (e.g., a target damping force) for controlling the damping force of the variable damper 7 of the suspension device 5 based on the instantaneous value input from the state estimation unit 17. Specifically, the control value calculation unit 20 calculates a suspension control value for improving the ride comfort of the vehicle based on, for example, bilinear optimal control, skyhook control, H∞ control, etc. The control value calculation unit 20 outputs a command current as a control signal to the damping force variable actuator 8 of the variable damper 7 based on the suspension control value. Note that the control value calculation unit 20 may calculate a suspension control value for improving not only the ride comfort of the vehicle but also the handling stability.

[0027] As shown in FIG. 2 , in this embodiment, the state estimation unit 17 estimates data related to suspension control for all four wheels (sprung speed and relative speed) by referencing behavior information, such as wheel speed, longitudinal acceleration, lateral acceleration, steering angle, and yaw rate, from the data constantly transmitted to the CAN 10. The estimated results are then transferred to the downstream control value calculation unit 20. At this time, the state estimation unit 17 estimates sensor data related to suspension control for all four wheels (vehicle state quantities). The relative speed estimation unit 18 and the sprung speed estimation unit 19 of the state estimation unit 17 are configured using neural networks. The weight parameters used in the neural network are determined in advance for the vehicle by machine learning. Therefore, the state estimation unit 17 can estimate the sprung speed and relative speed of all four wheels as vehicle state quantities, taking into account factors such as the rigidity characteristics of the entire vehicle. As a result, the control value calculation unit 20 controls the variable damper 7 based on the vehicle state quantities appropriately estimated by the state estimation unit 17. Therefore, the suspension device 5 can be controlled in accordance with the behavior of the vehicle, and the accuracy of the suspension control can be improved.

[0028] Next, the configuration and learning method of the neural networks of the relative velocity estimating section 18 and the sprung portion velocity estimating section 19 will be described with reference to FIGS.

[0029] The relative velocity estimation unit 18 and the sprung velocity estimation unit 19 are artificial intelligence (AI) and are configured by trained neural networks. Weight parameters and behavior information are input to the relative velocity estimation unit 18 and the sprung velocity estimation unit 19. The neural networks of the relative velocity estimation unit 18 and the sprung velocity estimation unit 19 are configured using weight parameters. As shown in FIGS. 3 and 4 , the relative velocity estimation unit 18 inputs behavior information into the neural network to estimate sensor data such as the relative velocities of four wheels and output the data. As shown in FIGS. 3 and 5 , the sprung velocity estimation unit 19 inputs behavior information into the neural network to estimate sensor data such as the sprung velocities of four wheels and output the data. Note that by regarding the vehicle body 2 as a rigid body, the sprung velocity of the remaining wheel can be geometrically estimated from the sprung velocities of three wheels. Therefore, the sprung velocity estimation unit 19 may estimate and output the sprung velocities of three wheels. 2, a state estimation unit 17 (relative speed estimation unit 18, sprung speed estimation unit 19) outputs the sprung speed and the relative speed as vehicle state quantities to a downstream control value calculation unit 20. The neural networks of the relative speed estimation unit 18 and the sprung speed estimation unit 19 are configured in the same manner as the neural network of the vehicle behavior estimation unit disclosed in, for example, Japanese Patent Application Laid-Open No. 2022-191913.

[0030] As an example, a specific configuration of the neural network of the relative velocity estimation unit 18 will be described with reference to Fig. 6. The neural network of the sprung velocity estimation unit 19 is configured in substantially the same manner as the neural network of the relative velocity estimation unit 18. Therefore, a description of the neural network of the sprung velocity estimation unit 19 will be omitted.

[0031] The neural network of the relative velocity estimation unit 18 is a three-layer hierarchical neural network in which elements of an input layer (number of elements i) 101, a hidden layer (number of elements j) 102, and an output layer (number of elements k) 103 are hierarchically connected. Each element of the input layer 101 is connected to each element of the hidden layer 102 by a weight W1ij (i = 1 to I, j = 1 to J), and each element of the hidden layer 102 is connected to each element of the output layer 103 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 by machine learning and stored in the weight parameter storage unit 14. Note that, in this example, a neural network with the simplest all-element-connected hidden layer 102 is shown, but this is not limiting. For example, a neural network with two or more hidden layers 102 may also be used.

[0032] As shown in Figure 4, time-series data such as longitudinal acceleration, lateral acceleration, steering angle, yaw rate, and wheel speeds for all four wheels contained in the CAN signal are input to the input layer 101 of the neural network. Instantaneous values ​​of the relative speeds of the suspension devices 5 for all four wheels of the vehicle are output to the output layer 103. The number of elements in the hidden layer 102 is generally determined by the number of elements in the input layer 101 and the output layer 103, and is set to a number that maximizes the accuracy of state estimation by the neural network. The number of elements in the output layer 103 is determined by the output specifications for state estimation.

[0033] The neural network machine learning is performed based on data on vehicle state quantities (e.g., sprung speed, unsprung speed) and behavior information data previously acquired by a data acquisition vehicle. The data acquisition vehicle has the same specifications as the vehicle in which the state estimation unit 17 (relative speed estimation unit 18, sprung speed estimation unit 19) is installed, and is equipped with various sensors (speed sensor, acceleration sensor, 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 the behavior information data. The weight parameters obtained as a result of learning are stored in the weight parameter memory unit 14.

[0034] In the present embodiment, the relative speed estimation unit 18 and the sprung speed estimation unit 19 perform machine learning to determine the correlation between the vehicle state quantity data and the behavior information data acquired by the data acquisition vehicle, but the present invention is not limited to this. For example, a vehicle model corresponding to the vehicle on which the relative speed estimation unit 18 and the sprung speed estimation unit 19 are installed may be constructed, and the relative speed estimation unit 18 and the sprung speed estimation unit 19 may perform machine learning based on data acquired by a simulation using this vehicle model.

[0035] Next, the estimation process of the state quantities by the neural networks of the relative velocity estimator 18 and the sprung velocity estimator 19 will be described with reference to FIGS. 3 to 5. FIG.

[0036] Fig. 3 shows a state estimation process using a neural network (artificial intelligence: AI) according to the first embodiment. Fig. 4 shows a relative velocity estimation process using a neural network according to the first embodiment. Fig. 5 shows a sprung velocity estimation process using a neural network according to the first embodiment.

[0037] In the first embodiment, wheel speed information for all four wheels is input to the neural network, so that the wheel speeds of the other wheels can be taken into account when estimating the state quantities of each wheel, thereby improving estimation accuracy. In this case, as shown in FIGS. 3 to 5 , the neural network that estimates the sprung velocity and relative velocity is divided into two parts, a relative velocity estimation unit 18 and a sprung velocity estimation unit 19, so that the neural networks can be operated individually at operation cycles appropriate for each state quantity. This improves ride comfort, while dividing the neural network into parts allows the processing load to be distributed. Furthermore, the state quantities to be output are not limited to the sprung velocity of each wheel; the neural network may also learn and output roll, pitch, and vertical velocity at any location.

[0038] Next, a specific example of the state estimation result by the sprung velocity estimator 19 will be described with reference to Fig. 7. Note that although the state estimation result by the sprung velocity estimator 19 will be described here, the state estimation result by the relative velocity estimator 18 can also provide the same effect as the above-described conventional technology.

[0039] The solid line in Fig. 7 indicates the sprung velocity estimated by the neural network (sprung velocity estimation unit 19) according to the first embodiment. On the other hand, the two-dot chain line in Fig. 7 indicates the actual measured value of the sprung velocity measured using a sensor. Furthermore, the dotted line in Fig. 7 indicates the sprung velocity estimated by the neural network according to the above-mentioned prior art.

[0040] In the above-described conventional technology, the sprung velocity and relative velocity of one wheel are estimated from wheel speed information for one wheel. Therefore, although the neural network receives an input signal from the wheel being estimated (e.g., the right front wheel), it does not receive input signals from other wheels (e.g., the right rear wheel, the left front wheel, and the left rear wheel). As a result, when vibrations occur in other wheels, the estimation device according to the above-described conventional technology cannot accurately estimate the effect of these vibrations on the wheel being estimated. Therefore, when traveling on a road surface different from the learned road surface spatial frequency, the timing of pitch and roll occurrences differs, resulting in a phase shift in the estimated sprung velocity, which tends to increase the difference between the estimated value and the sensor value.

[0041] In contrast, the sprung velocity estimator 19 according to the first embodiment estimates the state quantity of each wheel by inputting wheel speed information for all wheels (four wheels) into a neural network. Therefore, as shown in Fig. 7, the wheel speeds of other wheels can also be taken into account, so that even when pitch or roll occurs due to wheels other than the wheel being estimated, the influence on the wheel being estimated can be estimated with high accuracy, and the estimated value of the sprung velocity can be made closer to the sensor value.

[0042] Thus, the vehicle control device 1 according to this embodiment includes a variable damper 7 (relative displacement control device) that is provided between the vehicle body 2 and each of the plurality of wheels 3 of the vehicle and controls the relative displacement between the vehicle body 2 and the wheels 3, a vehicle state calculation unit 16 that is mounted on the vehicle and calculates the state of the vehicle, a relative speed estimation unit 18 that receives wheel speed information of each wheel 3 output from the vehicle state calculation unit 16 and estimates the relative speed of each wheel 3 using a neural network that has been trained to estimate the relative speed of each wheel 3 of the vehicle, a sprung speed estimation unit 19 that receives wheel speed information of each wheel 3 output from the vehicle state calculation unit 16 and estimates the sprung speed of each wheel 3 using a neural network that has been trained to estimate the sprung speed of each wheel 3 of the vehicle, and a control value calculation unit 20 (control unit) that controls the variable damper 7 based on outputs from the relative speed estimation unit 18 and the sprung speed estimation unit 19.

[0043] In the above-mentioned conventional technology, the size of each neural network is reduced and distributed across multiple neural networks, thereby widening the range of CPUs (Central Processing Units) that can be selected when implementing an ECU. However, in the above-mentioned conventional technology, because the state quantity is estimated from the wheel speed information of one wheel, there is a problem in that the estimation accuracy is low in relation to the effects of pitch and roll occurring in other wheels.

[0044] In contrast, in the first embodiment, a neural network is configured to estimate the state quantities of a plurality of wheels collectively, but the neural network is divided in terms of the operating cycle caused by the frequency components included in the state quantities to be estimated, etc. This makes it possible to improve the estimation accuracy while ensuring options for implementing the ECU (such as the operating cycle of the CPU and the number of cores).

[0045] As a result, by inputting information on multiple wheels into a single neural network and estimating the state quantities of the multiple wheels, estimation accuracy can be improved, resulting in improved ride comfort. Also, because the neural network is divided into a relative velocity estimation unit 18 and a sprung velocity estimation unit 19 according to the characteristics of the state quantities output by the neural network, it is possible to ensure estimation accuracy while also securing options for ECU implementation.

[0046] 1, 2, 8 to 10 show a second embodiment. The second embodiment is characterized in that the vehicle control device includes a relative speed estimation unit that estimates the relative speed of each wheel using a neural network that has been trained to estimate the relative speed of each wheel of the vehicle based on wheel speed information for three wheels excluding diagonal wheels from the wheel speed information of the wheels input from the vehicle state calculation unit, and a sprung speed estimation unit that estimates the sprung speed of each wheel using a neural network that has been trained to estimate the sprung speed of each wheel of the vehicle based on wheel speed information for three wheels excluding diagonal wheels from the wheel speed information of the wheels input from the vehicle state calculation unit. In the second embodiment, the same components as those in the first embodiment described above are designated by the same reference numerals, and their description will be omitted.

[0047] 1 and 2, a vehicle control device 21 according to the second embodiment is configured, like the vehicle control device 1 according to the first embodiment, by a suspension device 5 constituting a damping force generating device and a controller 22 constituting a vehicle control device. As shown in FIG. 2, the controller 22 is configured similarly to the controller 11 according to the first embodiment. Therefore, the controller 22 has a storage unit 12. The controller 22 also has a damper control unit 23, a weight parameter storage unit 14, and a data reading unit 15. The damper control unit 23 has a vehicle state calculation unit 16, a state estimation unit 24, and a control value calculation unit 20.

[0048] 8, various pieces of behavior information calculated based on the CAN signal are input to the state estimation unit 24 as input signals from the vehicle state calculation unit 16. At this time, the behavior information input to the state estimation unit 24 includes wheel speeds for all four wheels.

[0049] The state estimation unit 24 includes relative speed estimation units 25A to 25D that estimate the relative speed of each wheel, and sprung speed estimation units 26A to 26D that estimate the sprung speed of each wheel. The relative speed estimation unit 25A estimates the relative speed of the left front wheel. The relative speed estimation unit 25B estimates the relative speed of the right front wheel. The relative speed estimation unit 25C estimates the relative speed of the left rear wheel. The relative speed estimation unit 25D estimates the relative speed of the right rear wheel. The sprung speed estimation unit 26A estimates the sprung speed of the left front wheel. The sprung speed estimation unit 26B estimates the sprung speed of the right front wheel. The sprung speed estimation unit 26C estimates the sprung speed of the left rear wheel. The sprung speed estimation unit 26D estimates the sprung speed of the right rear wheel.

[0050] 9, relative speed estimation unit 25A receives wheel speed information for three wheels, excluding the wheel diagonal to the left front wheel (right rear wheel) that is the estimation wheel, out of the wheel speed information for four wheels received from vehicle state calculation unit 16. Relative speed estimation unit 25A receives the wheel speed information for three wheels output from vehicle state calculation unit 16, and estimates the relative speed of the left front wheel using a neural network that has been trained to estimate the relative speed of the left front wheel of the vehicle.

[0051] The relative speed estimation unit 25B receives wheel speed information for three wheels, excluding the wheel diagonal to the right front wheel (left rear wheel) that is the estimation wheel, out of the wheel speed information for four wheels received from the vehicle state calculation unit 16. The wheel speed information for three wheels output from the vehicle state calculation unit 16 is received by the relative speed estimation unit 25B, and the relative speed estimation unit 25B estimates the relative speed of the right front wheel using a neural network that has been trained to estimate the relative speed of the right front wheel of the vehicle.

[0052] The relative speed estimation unit 25C receives wheel speed information for three wheels, excluding the wheel diagonal to the left rear wheel (right front wheel) that is the estimation wheel, out of the wheel speed information for four wheels received from the vehicle state calculation unit 16. The wheel speed information for three wheels output from the vehicle state calculation unit 16 is received by the relative speed estimation unit 25C, and the relative speed of the left rear wheel is estimated using a neural network that has been trained to estimate the relative speed of the left rear wheel of the vehicle.

[0053] The relative speed estimation unit 25D receives wheel speed information for three wheels, excluding the wheel diagonal to the right rear wheel (left front wheel) that is the estimation wheel, out of the wheel speed information for four wheels received from the vehicle state calculation unit 16. The wheel speed information for three wheels output from the vehicle state calculation unit 16 is received by the relative speed estimation unit 25D, and the relative speed estimation unit 25D estimates the relative speed of the right rear wheel using a neural network that has been trained to estimate the relative speed of the right rear wheel of the vehicle.

[0054] 10 , the sprung velocity estimation unit 26A receives wheel speed information for three wheels excluding the wheel diagonal to the left front wheel (right rear wheel) that is the estimation wheel out of the wheel speed information for four wheels received from the vehicle state calculation unit 16. The wheel speed information for three wheels output from the vehicle state calculation unit 16 is received by the sprung velocity estimation unit 26A, and the sprung velocity estimation unit 26A estimates the sprung velocity of the left front wheel using a neural network that has been trained to estimate the sprung velocity of the left front wheel of the vehicle.

[0055] The sprung speed estimation unit 26B receives wheel speed information for three wheels excluding the wheel diagonal to the right front wheel (left rear wheel) that is the estimation wheel out of the wheel speed information for four wheels received from the vehicle state calculation unit 16. The wheel speed information for three wheels output from the vehicle state calculation unit 16 is received by the sprung speed estimation unit 26B, and the sprung speed estimation unit 26B estimates the sprung speed of the right front wheel using a neural network that has been trained to estimate the sprung speed of the right front wheel of the vehicle.

[0056] The sprung velocity estimation unit 26C receives wheel speed information for three wheels excluding the wheel diagonal to the left rear wheel (right front wheel) that is the estimation wheel out of the wheel speed information for four wheels received from the vehicle state calculation unit 16. The wheel speed information for three wheels output from the vehicle state calculation unit 16 is received by the sprung velocity estimation unit 26C, and the sprung velocity estimation unit 26C estimates the sprung velocity of the left rear wheel using a neural network that has been trained to estimate the sprung velocity of the left rear wheel of the vehicle.

[0057] The sprung speed estimation unit 26D receives wheel speed information for three wheels excluding the wheel diagonal to the right rear wheel (left front wheel) that is the estimation wheel out of the wheel speed information for four wheels received from the vehicle state calculation unit 16. The wheel speed information for three wheels output from the vehicle state calculation unit 16 is received by the sprung speed estimation unit 26D, and the sprung speed estimation unit 26D estimates the sprung speed of the right rear wheel using a neural network that has been trained to estimate the sprung speed of the vehicle's right rear wheel.

[0058] The weight parameters of the trained neural networks are stored in the weight parameter storage unit 14. Therefore, the neural networks of the relative speed estimation units 25A to 25D and the neural networks of the sprung part velocity estimation units 26A to 26D are configured using the weight parameters stored in the weight parameter storage unit 14. The neural networks of the relative speed estimation units 25A to 25D each output the relative speed of one wheel in response to input signals for three wheels from the vehicle state calculation unit 16. The neural networks of the sprung part velocity estimation units 26A to 26D each output the sprung part velocity of one wheel in response to input signals for three wheels from the vehicle state calculation unit 16.

[0059] Thus, the second embodiment can achieve substantially the same effects as the first embodiment. In the first embodiment, two neural networks are used to input the wheel speeds of four wheels and estimate the state quantities of the four wheels. In this case, the number of neural networks is reduced compared to the conventional method, but the processing load per neural network is increased. This makes it difficult to implement the system on a CPU with a low operating frequency. In contrast, in the second embodiment, the vehicle control device 21 includes relative speed estimation units 25A-25D that estimate the relative speed of each wheel using neural networks trained to estimate the relative speed of each wheel of the vehicle based on wheel speed information for three wheels excluding the diagonal wheels from the wheel speed information of the wheels input from the vehicle state calculation unit 16, and sprung speed estimation units 26A-26D that estimate the sprung speed of each wheel using neural networks trained to estimate the sprung speed of each wheel of the vehicle based on wheel speed information for three wheels excluding the diagonal wheels from the wheel speed information of the wheels input from the vehicle state calculation unit 16. That is, in the second embodiment, as shown in Figures 9 and 10, in order to estimate the effects of pitch and roll, the wheel speeds of three wheels excluding the wheel diagonal to the wheel to be estimated are input, and the state quantity of one wheel is estimated. Therefore, the estimation accuracy is improved compared to the above-mentioned conventional technology, while the processing load per neural network can be reduced compared to the first embodiment.

[0060] 1, 2, 11 to 13 show a third embodiment. The third embodiment is characterized in that the vehicle control device includes an unsprung speed estimation unit that receives wheel speed information for each wheel output from the vehicle state calculation unit and estimates the unsprung speeds of two wheels using a neural network that has been trained to estimate the unsprung speeds of two of the vehicle's wheels, a sprung speed estimation unit that receives wheel speed information for each wheel output from the vehicle state calculation unit and estimates the sprung speeds of three wheels using a neural network that has been trained to estimate the sprung speeds of three of the vehicle's wheels, and a control unit that controls the relative displacement suppression device based on outputs from the unsprung speed estimation unit and the sprung speed estimation unit. In the third embodiment, the same components as those in the first embodiment described above are designated by the same reference numerals, and their description will be omitted.

[0061] 1 and 2, a vehicle control device 31 according to the third embodiment is configured, similarly to the vehicle control device 1 according to the first embodiment, with a suspension device 5 constituting a damping force generating device and a controller 32 constituting a vehicle control device. The controller 32 is configured similarly to the controller 11 according to the first embodiment. Therefore, the controller 32 has a storage unit 12. The controller 32 also has a damper control unit 33, a weight parameter storage unit 14, and a data reading unit 15. The damper control unit 33 has a vehicle state calculation unit 16, a state estimation unit 34, and a control value calculation unit 20.

[0062] 11 , various types of behavior information calculated based on the CAN signal are input to the state estimation unit 34 as input signals from the vehicle state calculation unit 16. At this time, the behavior information input to the state estimation unit 34 includes wheel speeds for all four wheels.

[0063] The state estimation unit 34 includes an unsprung velocity estimation unit 35 that estimates the unsprung velocity of two of the vehicle's wheels (front wheels), and a sprung velocity estimation unit 37 that estimates the sprung velocity of three of the vehicle's wheels. In addition, the state estimation unit 34 further includes an unsprung velocity calculation unit 36 ​​that calculates the unsprung velocity of the rear wheels, a sprung velocity calculation unit 38 that calculates the sprung velocity of the remaining wheel, and a relative velocity calculation unit 39 that calculates the relative velocities of the four wheels.

[0064] The unsprung part velocity estimation unit 35 estimates the unsprung part velocity of the left front wheel and the right front wheel. The sprung part velocity estimation unit 37 estimates the sprung part velocity of the left front wheel, the right front wheel, and the left rear wheel, for example.

[0065] The unsprung velocity estimation unit 35 receives the wheel speed information for the four wheels output from the vehicle state calculation unit 16, and estimates the unsprung velocity of the left front wheel and the right front wheel using a neural network that has been trained to estimate the unsprung velocity of the left front wheel and the right front wheel of the vehicle.

[0066] The unsprung velocity calculation unit 36 ​​estimates the unsprung velocities of the two rear wheels (left rear wheel and right rear wheel) based on the unsprung velocities of the two front wheels (left front wheel and right front wheel) output from the unsprung velocity estimation unit 35. At this time, the unsprung velocity of the rear wheels is delayed by a time due to the vehicle speed, wheelbase, etc., relative to the unsprung velocity of the front wheels. For this reason, the unsprung velocity calculation unit 36 ​​calculates the unsprung velocity of the left rear wheel by performing delay processing for the wheelbase on the unsprung velocity of the left front wheel estimated by the unsprung velocity estimation unit 35. The unsprung velocity calculation unit 36 ​​calculates the unsprung velocity of the right rear wheel by performing delay processing for the wheelbase on the unsprung velocity of the right front wheel estimated by the unsprung velocity estimation unit 35. Note that when a deviation in trajectory occurs between the front wheels and the rear wheels as the vehicle turns, the unsprung velocity calculation unit 36 ​​may calculate the unsprung velocity taking this deviation in trajectory into account.

[0067] The sprung speed estimation unit 37 receives the wheel speed information for the four wheels output from the vehicle state calculation unit 16, and estimates the sprung speeds of the left front wheel, right front wheel, and left rear wheel using a neural network that has been trained to estimate the sprung speeds of three wheels of the vehicle (left front wheel, right front wheel, and left rear wheel).

[0068] The sprung velocity calculation unit 38 calculates the sprung velocity of the remaining right rear wheel from the sprung velocities of the three wheels (left front wheel, right front wheel, and left rear wheel) estimated by the sprung velocity estimation unit 37 by treating the vehicle body 2 as a rigid body.

[0069] The relative velocity calculation unit 39 calculates the relative velocities of the four wheels based on the difference between the unsprung velocities of the four wheels output from the unsprung velocity estimation unit 35 and the unsprung velocity calculation unit 36 ​​and the sprung velocities of the four wheels output from the sprung velocity estimation unit 37 and the sprung velocity calculation unit 38.

[0070] The control value calculation unit 20 controls the variable damper 7 based on the output (sprung velocity, relative velocity) from the state estimation unit 17. For this reason, in the third embodiment, the control value calculation unit 20 controls the variable damper 7 based on the outputs from the unsprung velocity estimation unit 35 and the sprung velocity estimation unit 37. The control value calculation unit 20 calculates a suspension control value (e.g., a target damping force) for controlling the damping force of the variable damper 7 of the suspension device 5, based on the instantaneous value input from the state estimation unit 17. The control value calculation unit 20 outputs a command current as a control signal to the damping force variable actuator 8 of the variable damper 7, based on the suspension control value.

[0071] Thus, the third embodiment can achieve substantially the same effects as the first embodiment. In the first and second embodiments, a total of eight state quantities, namely, the sprung velocities and relative velocities of four wheels, are estimated as the output of the neural network. However, by regarding the vehicle body 2 as a rigid body, it is possible to geometrically estimate the sprung velocity of the remaining wheel from the sprung velocities of three wheels. Furthermore, the relative velocities of four wheels can be calculated by estimating the unsprung velocity of the front wheels, then calculating the unsprung velocity of the rear wheels with a time delay, and finding the difference from the sprung velocity.

[0072] Therefore, in the third embodiment, the vehicle control device 31 has an unsprung speed estimation unit 35 that receives wheel speed information for each wheel output from the vehicle state calculation unit 16 and estimates the unsprung speeds of two wheels using a neural network that has been trained to estimate the unsprung speeds of two of the vehicle's wheels, an unsprung speed estimation unit 37 that receives wheel speed information for each wheel output from the vehicle state calculation unit and estimates the sprung speeds of three wheels using a neural network that has been trained to estimate the sprung speeds of three of the vehicle's wheels, and a control value calculation unit 20 (control unit) that controls the variable damper 7 based on outputs from the unsprung speed estimation unit 35 and the sprung speed estimation unit 37.

[0073] At this time, the unsprung velocity estimation unit 35 estimates the unsprung velocities of the two front wheels, the sprung velocity estimation unit 37 estimates the sprung velocities of the two front wheels and the left rear wheel, and the state estimation unit 34 calculates the relative velocities of the four wheels and the sprung velocities of the four wheels based on the unsprung velocities of the two wheels estimated by the unsprung velocity estimation unit 35 and the sprung velocities of the three wheels estimated by the sprung velocity estimation unit 37. For this reason, in the third embodiment, the processing load per neural network is lower than in the first embodiment, and the number of neural networks remains the same.

[0074] In the third embodiment, the sprung velocity estimator 37 estimates the sprung velocities of three wheels (left front wheel, right front wheel, left rear wheel) of the four wheels of the vehicle excluding the right rear wheel, but the present invention is not limited to this. The sprung velocity estimator 37 may estimate the sprung velocities of any three wheels of the four wheels of the vehicle, and may, for example, estimate the sprung velocities of three wheels excluding the left front wheel, three wheels excluding the right front wheel, or three wheels excluding the left rear wheel.

[0075] In the third embodiment, the unsprung speed estimator 35 estimates the unsprung speeds of the two front wheels, but the present invention is not limited to this. For example, when the vehicle is moving backward, the unsprung speed estimator 35 may estimate the unsprung speeds of the two rear wheels.

[0076] In the third embodiment, the control value calculation unit 20 (control unit) receives the relative velocities and sprung velocities for four wheels calculated based on the unsprung velocities for two wheels estimated by the unsprung velocity estimation unit 35 and the sprung velocities for three wheels estimated by the sprung velocity estimation unit 37. The control value calculation unit 20 then controls the variable damper 7 based on the relative velocities and sprung velocities for these four wheels, but the present invention is not limited to this. For example, the control value calculation unit (control unit) may receive the unsprung velocities for two wheels estimated by the unsprung velocity estimation unit 35 and the sprung velocities for three wheels estimated by the sprung velocity estimation unit 37. In this case, the control value calculation unit may internally calculate the relative velocities and sprung velocities for four wheels based on the unsprung velocities for two wheels and the sprung velocities for three wheels, and control the variable damper 7 based on the relative velocities and sprung velocities for these four wheels.

[0077] In the first embodiment, the controller 11 acquires vehicle behavior information (vehicle state) including wheel speed through the CAN 10, but the present invention is not limited to this. The controller 11 may, for example, directly acquire detection values ​​from various sensors. Alternatively, various sensors that detect longitudinal acceleration (longitudinal G), lateral acceleration (lateral G), steering angle, yaw rate, wheel speed, etc. may calculate the vehicle state and input the calculated value to the controller 11. In this case, the various sensors constitute a vehicle state calculation unit. Alternatively, the controller 11 may acquire behavior information from another controller, etc. This configuration can be applied to the second and third embodiments.

[0078] In the above-described embodiments, the force generating mechanism is a variable damper 7 made up 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 mechanism. In the above-described embodiments, the force generating mechanism that generates an adjustable force between the vehicle body 2 side and the wheel 3 side is configured by a variable damper 7 made up of a damping force adjustable hydraulic shock absorber. However, the present invention is not limited to this, and the force generating mechanism may be configured by an air suspension, a stabilizer (kinesus), an electromagnetic suspension, or the like, in addition to a hydraulic shock absorber.

[0079] In the above embodiments, the vehicle behavior control device is used for a four-wheeled automobile. However, the present invention is not limited to this, and can also be applied to, for example, a work vehicle, a transport vehicle such as a truck or a bus.

[0080] This application claims priority to Japanese Patent Application No. 2023-198141, filed November 22, 2023. The entire disclosure of Japanese Patent Application No. 2023-198141, filed November 22, 2023, including the specification, claims, drawings, and abstract, is incorporated herein by reference in its entirety.

[0081] 1, 21, 31: vehicle control device, 2: vehicle body, 3: wheel, 5: suspension device, 7: adjustable damping force shock absorber (variable damper), 8: variable damping force actuator, 11, 22, 32: controller, 16: vehicle state calculation unit, 17, 24, 34: state estimation unit, 18, 25A to 25D: relative velocity estimation unit, 19, 26A to 26D, 37: sprung velocity estimation unit, 20: control value calculation unit (control unit), 35: unsprung velocity estimation unit, 36: unsprung velocity calculation unit, 38: sprung velocity calculation unit, 39: relative velocity calculation unit

Claims

1. A vehicle control device comprising: a relative displacement control device provided between a vehicle body and each of a plurality of wheels, the relative displacement control device controlling the relative displacement between the vehicle body and the wheels; a vehicle state calculation unit mounted on the vehicle and calculating a state of the vehicle; a relative speed estimation unit receiving wheel speed information of each wheel output from the vehicle state calculation unit and estimating the relative speed of each wheel using a neural network that has been trained to estimate the relative speed of each wheel of the vehicle; a sprung velocity estimation unit receiving wheel speed information of each wheel output from the vehicle state calculation unit and estimating 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; and a control unit that controls a relative displacement suppression device based on outputs from the relative speed estimation unit and the sprung velocity estimation unit.

2. A vehicle control device comprising: a relative displacement control device provided between a vehicle body and each of a plurality of wheels, the relative displacement control device controlling the relative displacement between the vehicle body and the wheels; a vehicle state calculation unit mounted on the vehicle and calculating a state of the vehicle; a relative speed estimation unit estimating the relative speed of each wheel using a neural network that has been trained to estimate the relative speed of each wheel of the vehicle based on wheel speed information of the wheels for three wheels excluding diagonal wheels from the wheel speed information of the wheels input from the vehicle state calculation unit; a sprung speed estimation unit estimating the sprung speed of each wheel using a neural network that has been trained to estimate the sprung speed of each wheel of the vehicle based on wheel speed information of the wheels for three wheels excluding diagonal wheels from the wheel speed information of the wheels input from the vehicle state calculation unit; and a control unit that controls a relative displacement suppression device based on outputs from the relative speed estimation unit and the sprung speed estimation unit.

3. A vehicle control device comprising: a relative displacement control device provided between a vehicle body and each of a plurality of wheels, the relative displacement control device controlling the relative displacement between the vehicle body and the wheels; a vehicle state calculation unit mounted on the vehicle and calculating the state of the vehicle; an unsprung velocity estimation unit receiving wheel speed information of each wheel output from the vehicle state calculation unit and estimating the unsprung velocities of the two wheels using a neural network that has been trained to estimate the unsprung velocities of the two wheels; a sprung velocity estimation unit receiving wheel speed information of each wheel output from the vehicle state calculation unit and estimating the sprung velocities of the three wheels using a neural network that has been trained to estimate the sprung velocities of the three wheels; and a control unit that controls a relative displacement suppression device based on outputs from the unsprung velocity estimation unit and the sprung velocity estimation unit.

4. A vehicle control device as described in claim 3, wherein the unsprung velocity estimation unit estimates the unsprung velocity of the front two wheels, the unsprung velocity estimation unit estimates the sprung velocity of the front two wheels and the left or right rear wheel, and the vehicle control device calculates the relative velocities and the sprung velocities of the four wheels based on the unsprung velocities of the two wheels estimated by the unsprung velocity estimation unit and the sprung velocity of the three wheels estimated by the unsprung velocity estimation unit.

Images