Braking force distribution method and device for chassis control of unmanned vehicle

Abstract

The embodiment of the invention discloses a braking force distribution method and device for chassis control of an unmanned vehicle. The method comprises the steps that according to an acceleration instruction, a vehicle dynamics model and a PI controller are applied to obtain required torque; determining whether the vehicle is in a driving mode or a braking mode according to the required torque; and if it is determined that the vehicle is in a braking mode, according to the required torque, under the criterion of feeding back electric energy as much as possible, an electric braking torque instruction and an air braking reference value are obtained, the air braking reference value is adjusted based on control parameters adjusted by a neural network, and an air braking instruction is obtained. By means of the mode, on the premise that the maximum electric braking is guaranteed, the rapidity and accuracy of braking distribution and composite braking are improved, and meanwhile the accuracy and real-time performance of pneumatic braking are improved.

Description

Technical Field

[0001] The embodiments of the present invention relate to the field of vehicle braking technology, and more particularly to a method and device for distributing braking force for chassis control of an unmanned vehicle. Background Art

[0002] Autonomous vehicles can significantly reduce traffic accidents caused by human factors such as fatigue and drunk driving, improve public transportation safety, and alleviate urban congestion. They have seen significant development in recent years. Autonomous vehicles utilize advanced perception and decision-making control technologies to calculate optimal driving instructions and then control the vehicle chassis to follow these instructions. Therefore, the functional implementation of autonomous vehicles is divided into three layers: perception, decision, and execution. The drive-by-wire chassis is the core component of the execution layer. The drive-by-wire chassis receives signal instructions such as distance, speed, acceleration, deceleration, and steering commands from the decision layer via a digital communication bus. Through precise control logic and algorithms, the vehicle's actual motion tracks the received instructions. The chassis' acceleration and braking control performance is crucial for autonomous driving trajectory tracking and directly determines the implementation of advanced autonomous driving functions such as following, stopping, lane changing, overtaking, and parking. Furthermore, braking performance directly impacts driving safety. Appropriate braking control strategies can also provide increased energy return, extending the vehicle's range.

[0003] In a purely electric autonomous vehicle, driving force is provided by the power output of a lithium-ion battery, while braking force is a combination of electrical energy feedback and the friction of the brake pads. Chassis control involves using algorithms to adjust the appropriate driving and braking forces based on the current state of motion, thereby tracking the trajectory of the decision-making layer. Trajectory tracking requires that distance, speed, and acceleration / deceleration be controlled within certain accuracy limits. For example, distance accuracy must be ±2m, speed accuracy must be ±1km / h, and acceleration / deceleration accuracy must be ±0.05m / s². Feedback control or optimization control algorithms are typically used to convert distance and speed tracking into acceleration / deceleration commands. This means that distance and speed deviations are compensated for by appropriately adjusting the acceleration / deceleration. Therefore, acceleration / deceleration tracking is a core function of chassis control.

[0004] The acceleration / deceleration control algorithm converts acceleration / deceleration commands issued by the upper layer into motor torque commands or brake control values ​​for the electronic brake system (EBS), thereby controlling the execution of the mechanical mechanism. The acceleration and deceleration tracking control algorithms share common design elements. Both convert acceleration / deceleration commands into torque control commands. The difference is that with acceleration control, the torque command is executed solely by the motor; whereas with deceleration control, the torque command must be shared by both the motor's regenerative braking and the air brake. In comparison, deceleration control is more complex. Therefore, braking control is the key and most challenging aspect of chassis control.

[0005] The key to chassis braking control is designing a rational braking force distribution algorithm. With the goals of safe braking, fast and accurate tracking, and maximum regenerative energy, this algorithm distributes braking force to the motor and air brake actuators by combining various parameters, including vehicle speed, deceleration command, the inherent characteristics of the air brake, and the motor's regenerative braking capability.

[0006] Currently, there are two types of methods for braking force distribution. One is rule-based, where allocation rules are designed based on parameters such as vehicle speed, braking torque command, maximum motor braking torque, and air brake delay time. This method is highly practical, and the rules can be continuously updated based on test experience. However, this method is not universal across different vehicle models, and the large number of rules complicates the process. The other is optimization-based, where an optimization objective is constructed, constraints are used as constraints, and the optimal solution is automatically determined. This method is highly versatile and produces the most reasonable braking distribution. However, it is computationally intensive and relies on an accurate model. Summary of the Invention

[0007] In view of the above problems, embodiments of the present invention provide a method and device for distributing braking force for chassis control of an unmanned vehicle, which overcomes the above problems or at least partially solves the above problems.

[0008] According to one aspect of an embodiment of the present invention, a braking force distribution method for chassis control of an unmanned vehicle is provided, the method comprising: applying a vehicle dynamics model and a PI controller to obtain a required torque according to an acceleration instruction; determining whether the vehicle is in a driving mode or a braking mode according to the required torque; if the vehicle is determined to be in a braking mode, obtaining an electric braking torque instruction and an air braking reference value according to the required torque and under the criterion of feeding back as much electrical energy as possible, and adjusting the air braking reference value based on a control parameter adjusted by a neural network to obtain an air braking instruction.

[0009] Optionally, the applying of the vehicle dynamics model and the PI controller to obtain the required torque according to the acceleration instruction includes: applying the vehicle dynamics model to calculate the torque feedforward value according to the acceleration instruction; applying the PI controller to calculate the torque percentage according to the difference between the acceleration instruction and the actual acceleration; calculating the product of the torque percentage and the maximum driving / braking torque to obtain a torque compensation value; and applying the torque compensation value to compensate the torque feedforward value to obtain the required torque.

[0010] Optionally, determining whether the vehicle is in driving mode or braking mode based on the required torque includes: if the required torque is output as 1 after the hysteresis loop, determining that the vehicle is in driving mode; if the required torque is output as 0 after the hysteresis loop, determining that the vehicle is in braking mode.

[0011] Optionally, the electric braking torque instruction and the air braking reference value are obtained based on the required torque and under the criterion of feeding back as much electric energy as possible, including: selecting the maximum value from the required torque and the maximum braking torque of the motor as the electric braking torque instruction; if the required torque exceeds the maximum braking torque of the motor, calculating the first torque difference between the required torque and the maximum braking torque; and obtaining the air braking reference value based on the required torque, the first torque difference and the maximum braking power of the motor.

[0012] Optionally, obtaining the air brake reference value based on the required torque, the first torque difference, and the maximum braking power of the motor includes: calculating the difference between the required torque and the torque corresponding to the maximum braking power of the motor to obtain a second torque difference; and calculating the air brake reference value based on the smaller value of the first torque difference and the second torque difference, and the calculation relationship applied is as follows:

[0013]

[0014] Among them, α airref is the air brake reference value, α cmd is the acceleration command, ΔT q is the torque difference, that is, the smaller value of the first torque difference and the second torque difference, T ref is the required torque.

[0015] Optionally, the control parameters adjusted based on the neural network adjust the air brake reference value to obtain an air brake instruction, including: obtaining the difference between the acceleration instruction and the acceleration feedback value to obtain the acceleration difference; using the acceleration difference and the air brake compensation value output by the PI controller as input, and applying the neural network to adjust the parameters of the PI controller online with the goal of minimizing the mean square error of the acceleration difference; applying the PI controller to obtain the air brake compensation value according to the acceleration difference and the adjusted parameters; and applying the air brake compensation value to compensate the air brake reference value to obtain an air brake instruction.

[0016] Optionally, the method further includes: if it is determined that the vehicle is in a driving mode, controlling the required torque to be directly output to the motor for motor driving.

[0017] Based on the same inventive concept, a braking force distribution device for chassis control of an unmanned vehicle is provided, including: a demand torque acquisition unit, used to apply a vehicle dynamics model and a PI controller to obtain the demand torque according to an acceleration instruction; a mode judgment unit, used to determine whether the vehicle is in a driving mode or a braking mode according to the demand torque; a braking force distribution unit, used to obtain an electric braking torque instruction and an air braking reference value according to the demand torque and under the criterion of feeding back as much electric energy as possible if the vehicle is determined to be in a braking mode, and adjust the air braking reference value based on the control parameters adjusted by the neural network to obtain an air braking instruction.

[0018] Based on the same inventive concept, an embodiment of the present invention further proposes an electronic device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor implements the aforementioned method when executing the program.

[0019] Based on the same inventive concept, an embodiment of the present invention further proposes a computer storage medium, in which at least one executable instruction is stored, and the executable instruction enables a processor to execute the aforementioned method.

[0020] The embodiment of the present invention obtains the required torque by applying a vehicle dynamics model and a PI controller according to an acceleration instruction; determines whether the vehicle is in a driving mode or a braking mode according to the required torque; if the vehicle is determined to be in a braking mode, obtains an electric braking torque instruction and an air braking reference value according to the required torque and under the principle of feeding back as much electric energy as possible, and adjusts the air braking reference value based on control parameters adjusted by a neural network to obtain an air braking instruction. This can improve the speed and accuracy of brake distribution and compound braking while ensuring maximum electric braking, and by applying a neural network for online parameter adjustment, simultaneously improve the accuracy and real-time performance of air braking.

[0021] The above description is only an overview of the technical solutions of the embodiments of the present invention. In order to more clearly understand the technical means of the embodiments of the present invention, they can be implemented in accordance with the contents of the specification. In order to make the above and other purposes, features and advantages of the embodiments of the present invention more obvious and easy to understand, the specific implementation methods of the present invention are specifically listed below. BRIEF DESCRIPTION OF THE DRAWINGS

[0022] Various other advantages and benefits will become apparent to those skilled in the art upon reading the detailed description of the preferred embodiment below. The accompanying drawings are for illustration purposes only and are not to be considered as limiting the present invention. The same reference symbols are used throughout the drawings to represent the same components. In the drawings:

[0023] Figure 1 A schematic flow chart of a method for distributing braking force for chassis control of an unmanned vehicle provided by an embodiment of the present invention is shown;

[0024] Figure 2 An overall block diagram of braking force distribution for chassis control of an unmanned vehicle according to an embodiment of the present invention is shown;

[0025] Figure 3 A schematic diagram of an acceleration / deceleration control loop according to an embodiment of the present invention is shown;

[0026] Figure 4 A schematic diagram of reset logic in a PI controller according to an embodiment of the present invention is shown;

[0027] Figure 5 A schematic diagram of static start compensation of a PI controller according to an embodiment of the present invention is shown;

[0028] Figure 6 A schematic structural diagram of a braking force distribution device for chassis control of an unmanned vehicle provided by an embodiment of the present invention is shown;

[0029] Figure 7 A schematic diagram of an electronic device in an embodiment of the present invention is shown. DETAILED DESCRIPTION

[0030] Exemplary embodiments of the present invention will be described in more detail below with reference to the accompanying drawings. Although exemplary embodiments of the present invention are shown in the accompanying drawings, it should be understood that the present invention can be implemented in various forms and should not be limited by the embodiments set forth herein. Rather, these embodiments are provided to enable a more thorough understanding of the present invention and to fully convey the scope of the present invention to those skilled in the art.

[0031] Figure 1 FIG2 shows a flow chart of a method for distributing braking force for chassis control of an unmanned vehicle provided by an embodiment of the present invention. Figure 1As shown, the braking force distribution method for chassis control of an unmanned vehicle is applied to a server, including:

[0032] Step S11: Applying the vehicle dynamics model and the PI controller to obtain the required torque according to the acceleration command.

[0033] In an embodiment of the present invention, the overall block diagram of the adaptive braking force distribution of the unmanned vehicle chassis control is as follows: Figure 2 As shown in the figure, the system consists of two main parts: the acceleration / deceleration loop and the braking force distribution. First, the acceleration command passes through the acceleration / deceleration loop and is converted into a required torque, which serves as the reference torque input. Then, based on the vehicle and motor conditions, the reference torque is adaptively distributed into the torque command for the motor and the deceleration command for the air brake.

[0034] In step S11, optionally, a vehicle dynamics model is applied to calculate a torque feedforward value based on an acceleration command; a PI controller is applied to calculate a torque percentage based on the difference between the acceleration command and the actual acceleration; the product of the torque percentage and the maximum driving / braking torque is calculated to obtain a torque compensation value; the torque compensation value is applied to compensate the torque feedforward value to obtain the required torque. The specific block diagram is as follows: Figure 3 As shown, the input of the acceleration / deceleration loop is the instruction issued by the upper layer, and the output is the total required torque T ref The torque feedforward value T is calculated using the vehicle dynamics model. ff The parameters of this vehicle dynamics model do not need to be accurate; only the parameters of an unloaded vehicle need to be configured. The deviation of the vehicle dynamics model is compensated by the PI controller. Figure 3 The maximum driving / braking torque in is the actual torque capacity of the motor T nrkmax , this value is given by the motor controller in real time. Figure 3 The compensation part uses the output method of torque percentage, so that the parameters can be calibrated in a unified manner for different models.

[0035] When the vehicle is stationary and the deceleration command is in effect, or when the driving mode and the braking mode are switched, the PI controller must be reset to clear the integral. The reset logic of the PI controller is as follows: Figure 4 As shown in Figure 1, the deceleration command is a deceleration command ≤ 0, and the reset logic value is transmitted to the PI controller for reset at an appropriate time. The reset logic in the PI controller can improve the response speed.

[0036] When the vehicle is stationary, it needs a larger torque to start. In order to eliminate the overshoot of the control amount under the small acceleration command, a certain compensation is given at the moment of starting. The static start compensation of the PI controller is as follows: Figure 5As shown in the figure, compensation is applied when the vehicle is stationary and the acceleration command is small (<0.5). The compensation amount is a calibrated value that is customized for each vehicle model, typically 4%. The stationary start compensation value is directly transmitted to the PI controller for compensation when necessary.

[0037] Step S12: determining whether the vehicle is in a driving mode or a braking mode according to the required torque.

[0038] In the embodiment of the present invention, according to the required torque, that is, the torque reference value T ref To determine the current driving / braking mode of the vehicle. If the required torque output after the hysteresis loop is 1, the vehicle is determined to be in driving mode; if the required torque output after the hysteresis loop is 0, the vehicle is determined to be in braking mode. The critical value of the hysteresis loop is set to ±K smooth ·T max , T max is the maximum torque, the empirical value K smooth =2.85%. By adjusting the hysteresis parameter K smooth , which can eliminate the jitter of switching between the two modes.

[0039] Step S13: If it is determined that the vehicle is in braking mode, the electric brake torque command and the air brake reference value are obtained according to the required torque and under the principle of feeding back as much electric energy as possible, and the air brake reference value is adjusted based on the control parameters adjusted by the neural network to obtain the air brake command.

[0040] In the embodiment of the present invention, if it is determined that the vehicle is in the driving mode, the control directly outputs the required torque T ref To the motor for motor driving, that is, the motor torque command T motcmd =T ref , at this time the air brake command α aircmd Output invalid value, that is, the air brake does not respond.

[0041] If the vehicle is determined to be in braking mode, the motor's regenerative braking torque must be calculated based on the torque reference value and motor status, maximizing the amount of regenerative electrical energy. If electric braking is insufficient, air braking is used to supplement, and the air brake deceleration is calculated and sent to the air brake actuator. This means distributing the braking torque, outputting motor braking torque commands and air brake deceleration commands. Braking force distribution must overcome two challenges: 1) jitter from switching between the two modes; and 2) adjustment lag and deviation due to the air brake actuator's large time delay and strong nonlinearity.

[0042] In the embodiment of the present invention, in order to recover the braking energy to the maximum extent, electric braking is preferred. Optionally, the maximum value is first selected from the required torque and the maximum braking torque of the motor as the electric braking torque command, that is, Tmotcmd =max(T ref ,T brkmax ). Among them, T brkmax It is the maximum braking torque of the motor, which is a negative value.

[0043] If the required torque exceeds the maximum braking torque of the motor, that is, T ref <T brkmax , then the part of the braking force other than that borne by the motor is allocated to the air brake. At this time, the first torque difference ΔT between the required torque and the maximum braking torque is calculated q1 , ΔT q1 =T ref -T brkmax Then, the air brake reference value is obtained based on the required torque, the first torque difference and the maximum braking power of the motor. Optionally, the difference between the required torque and the torque corresponding to the maximum braking power of the motor is calculated based on the maximum braking power limit of the motor to obtain the second torque difference ΔT q2 , R w is the radius of the vehicle tire, V is the vehicle speed; during braking, the torque difference is a negative value. The air brake reference value α is calculated based on the smaller value of the first torque difference and the second torque difference. airref . According to the physical formula

[0044] T q =F·L≈M·α·L

[0045] Where, T q is the torque, F is the force, L is the rotation radius, M is the mass, and a is the acceleration. It can be seen that α and T q are equivalent and approximately proportional. According to T q To be proportional to α, calculate the air brake reference value α airref The calculation relationship applied is as follows:

[0046]

[0047] Among them, α airref is the air brake reference value, α cmd is the acceleration instruction, T ref is the required torque, ΔT q is the torque difference, that is, the smaller value of the first torque difference and the second torque difference, ΔT q =min(ΔT q1 ,ΔT q2 ).

[0048] Due to T q =F·L≈M·α·L is an approximate equation, so αairref There will definitely be deviations. In order to compensate for the deviations, we need to airref This feedforward value is used as the air brake's feedforward value, supplemented by feedback compensation. This feedforward-plus-compensation control method allows for rapid tracking of deceleration commands while maintaining maximum energy recovery efficiency. Furthermore, compensation control compensates for the air brake's inherent control deviations and the effects of slope, achieving higher control accuracy across various road sections.

[0049] In the embodiment of the present invention, in order to overcome the delay characteristics, nonlinear characteristics, and the influence of different slopes and road factors during the execution of the air brake, continue to refer to Figure 2 , using a neural network to perform online parameter adjustment on the air brake compensation controller. Optionally, first obtain the difference between the acceleration command and the acceleration feedback value to obtain the acceleration difference e(k); the acceleration difference e(k) and the air brake compensation value α output by the PI controller are used to calculate the air brake compensation value α. airfb The invention adopts a novel method for adjusting the parameters of the PI controller online by applying a neural network with the goal of minimizing the mean square error of the acceleration difference; the PI controller is applied to obtain an air brake compensation value according to the acceleration difference and the adjusted parameters; the air brake reference value is compensated by applying the air brake compensation value to obtain an air brake instruction.

[0050] In the embodiment of the present invention, the input of the neural network is e(k), u(k), and the output is K p (k), K i (k). The structure consists of an input layer, a hidden layer, and an output layer. The number of neural units in each layer is 4, 4, and 2, respectively. The activation function of the hidden layer uses the sigmoid function. The activation function of the output layer uses the ReLU function to avoid negative output values. The entire neural network is trained and adjusted online through backpropagation. The specific signals of the input layer are as follows:

[0051]

[0052] Let the input of the hidden layer be S j , the calculation relationship is as follows:

[0053]

[0054] Among them, ω ij is the weight between the input layer and the hidden layer. i After the sigmoid activation function, the output h of the hidden layer is obtained i .

[0055] Let the input of the output layer be Z j , the calculation relationship is as follows:

[0056]

[0057] in, is the weight between the hidden layer and the output layer. i After the ReLU activation function, the output o of the hidden layer is obtained j .

[0058] The entire neural network can be expressed as a general function as shown below:

[0059]

[0060] Minimize the mean square error E(k) = 1 / 2e(k) by gradient descent 2 , and then adopt the back propagation method, starting from the output layer, updating the weights layer by layer to achieve the purpose of training the neural network.

[0061] For the output layer, the mean square error E(k) is the weight The derivative of is as follows:

[0062]

[0063] Among them, the gradient function of each part is specifically:

[0064]

[0065] For the hidden layer, the mean square error E(k) is the weight ω ij The derivative of is as follows

[0066]

[0067] Among them, i∈[1,4], j∈[1,2], the gradient function of each part is specifically

[0068]

[0069] Combining the above formulas, the weight update of the neural network is shown as follows:

[0070]

[0071] Among them, η is the learning rate, which is set to 0.01.

[0072] The final output PI controller parameters are shown in the following formula:

[0073]

[0074] Taking the acceleration difference e(k) as input, the PI controller is applied based on the adjusted parameters to obtain the air brake compensation value α airfb , calculate the air brake compensation value α airfband air brake reference value α airref The sum of the air brake command α aircmd , the air brake instruction α aircmd The data is then transmitted to the corresponding air brake actuator to execute the air brake. The PI controller parameters are adjusted online in real time through the neural network to improve the accuracy and speed of the air brake's deceleration adjustment.

[0075] The braking force distribution method for chassis control of unmanned vehicles according to the present invention can be applied to autonomous buses. Several practical challenges exist in braking distribution and control for autonomous buses. 1) The mass of a bus varies significantly between empty and fully loaded, making it difficult to obtain accurate values ​​in real time. 2) The response accuracy and latency of air brakes are inferior to those of passenger cars, especially during low-speed braking, which increases the difficulty of brake distribution control. 3) Autonomous driving requires high response accuracy and latency. 4) Accurate speed stabilization and stopping on slopes must be achieved regardless of slope information. The braking force distribution method for chassis control of unmanned vehicles according to the present invention employs a compensation controller and a static start compensation strategy to address the issue of inaccurate dynamic models, significantly reducing overshoot in acceleration / deceleration control and improving steady-state accuracy. Furthermore, to address the poor accuracy and latency of air brake components in bus chassis, a neural network online parameter adjustment method is designed to improve the accuracy and real-time performance of air brakes.

[0076] In summary, the braking force distribution method for chassis control of an unmanned vehicle in an embodiment of the present invention obtains the required torque by applying a vehicle dynamics model and a PI controller according to an acceleration instruction; determines whether the vehicle is in a driving mode or a braking mode according to the required torque; if the vehicle is determined to be in a braking mode, obtains an electric braking torque instruction and an air braking reference value according to the required torque and under the criterion of feeding back as much electric energy as possible, adjusts the air braking reference value based on the control parameters adjusted by the neural network, and obtains an air braking instruction; based on the physical laws of torque and acceleration, a method combining a deceleration feedforward control law and compensation is designed, which improves the speed and accuracy of braking distribution and compound braking under the premise of ensuring maximum electric braking; and by applying a neural network for online parameter adjustment, simultaneously improves the accuracy and real-time performance of air braking.

[0077] The foregoing description is of specific embodiments of the present invention. In some cases, the actions or steps described in the embodiments of the present invention may be performed in an order different from that shown in the embodiments and still achieve the desired results. In addition, the processes depicted in the accompanying drawings do not necessarily require the specific order or sequential order shown to achieve the desired results. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0078] Based on the same concept, an embodiment of the present invention also provides a braking force distribution device for chassis control of an unmanned vehicle. Figure 6 As shown, the braking force distribution device for the chassis control of the unmanned vehicle includes: a required torque acquisition unit, a mode judgment unit and a braking force distribution unit.

[0079] a demand torque acquisition unit, for obtaining the demand torque by applying a vehicle dynamics model and a PI controller according to an acceleration command;

[0080] a mode determination unit, configured to determine whether the vehicle is in a driving mode or a braking mode according to the required torque;

[0081] The braking force distribution unit is used to obtain an electric braking torque command and an air braking reference value based on the required torque and the principle of feeding back as much electrical energy as possible if the vehicle is determined to be in braking mode, and adjust the air braking reference value based on the control parameters adjusted by the neural network to obtain an air braking command.

[0082] For the convenience of description, the above devices are described as being divided into various modules according to their functions. Of course, when implementing the embodiments of the present invention, the functions of each module can be implemented in the same or multiple software and / or hardware.

[0083] The apparatus of the above embodiment is applied to the corresponding method of the above embodiment and has the beneficial effects of the corresponding method embodiment, which will not be described in detail here.

[0084] Based on the same inventive concept, an embodiment of the present invention further provides an electronic device, which includes a memory, a processor, and a computer program stored in the memory and runnable on the processor, wherein when the processor executes the program, the method described in any one of the above embodiments is implemented.

[0085] An embodiment of the present invention provides a non-volatile computer storage medium, wherein the computer storage medium stores at least one executable instruction, and the computer executable instruction can execute the method described in any one of the above embodiments.

[0086] Figure 7 : A more specific hardware structure diagram of an electronic device provided in this embodiment is shown. The device may include: a processor 701, a memory 702, an input / output interface 703, a communication interface 704, and a bus 705. The processor 701, the memory 702, the input / output interface 703, and the communication interface 704 are communicatively connected to each other within the device via the bus 705.

[0087] The processor 701 can be implemented as a general-purpose CPU (Central Processing Unit), a microprocessor, an application-specific integrated circuit (ASIC), or one or more integrated circuits, and is used to execute relevant programs to implement the technical solutions provided by the method embodiments of the present invention.

[0088] The memory 702 can be implemented in the form of ROM (Read Only Memory), RAM (Random Access Memory), static storage devices, dynamic storage devices, etc. The memory 702 can store an operating system and other application programs. When the technical solutions provided by the method embodiments of the present invention are implemented through software or firmware, the relevant program codes are stored in the memory 702 and called and executed by the processor 701.

[0089] The input / output interface 703 is used to connect to the input / output module to realize information input and output. The input / output module can be configured as a component in the device (not shown in the figure) or can be externally connected to the device to provide corresponding functions. Input devices may include a keyboard, mouse, touch screen, microphone, various sensors, etc., and output devices may include a display, speaker, vibrator, indicator light, etc.

[0090] The communication interface 704 is used to connect to a communication module (not shown) to enable communication between the device and other devices. The communication module can communicate via a wired method (such as USB, network cable, etc.) or a wireless method (such as mobile network, WIFI, Bluetooth, etc.).

[0091] The bus 705 comprises a pathway for transmitting information between the various components of the device (eg, the processor 701 , the memory 702 , the input / output interface 703 , and the communication interface 704 ).

[0092] It should be noted that although the above device only shows the processor 701, the memory 702, the input / output interface 703, the communication interface 704, and the bus 705, in a specific implementation, the device may also include other components necessary for normal operation. In addition, those skilled in the art will understand that the above device may only include the components necessary to implement the embodiments of the present invention, and does not necessarily include all the components shown in the figure.

[0093] Those skilled in the art should understand that the discussion of any of the above embodiments is merely illustrative and is not intended to imply that the scope of the present disclosure is limited to these examples. Based on the concept of the present disclosure, the technical features in the above embodiments or different embodiments may be combined, the steps may be implemented in any order, and there are many other variations of the different aspects of the embodiments of the present invention as described above, which are not provided in detail for the sake of simplicity.

[0094] This application is intended to cover all such substitutions, modifications, and variations that fall within the broad scope of all embodiments. Therefore, any omissions, modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the embodiments of the present invention should be included in the scope of protection of this disclosure.

Claims

1. A braking force distribution method for chassis control of an unmanned vehicle, characterized in that: The method comprises: Apply the vehicle dynamics model and PI controller to obtain the required torque according to the acceleration command; determining whether the vehicle is in a driving mode or a braking mode according to the required torque; If it is determined that the vehicle is in braking mode, the electric brake torque command and the air brake reference value are obtained according to the required torque and under the principle of feeding back as much electrical energy as possible, and the air brake reference value is adjusted based on the control parameters adjusted by the neural network to obtain the air brake command.

2. The method according to claim 1, characterized in that The step of applying the vehicle dynamics model and the PI controller to obtain the required torque according to the acceleration command includes: Calculate torque feedforward value based on acceleration command using vehicle dynamics model; Calculating the torque percentage using a PI controller according to the difference between the acceleration command and the actual acceleration; Calculating the product of the torque percentage and the maximum driving / braking torque to obtain a torque compensation value; The torque compensation value is applied to compensate the torque feed-forward value to obtain the required torque.

3. The method according to claim 1, characterized in that: The step of determining whether the vehicle is in a driving mode or a braking mode according to the required torque includes: If the required torque is output as 1 after the hysteresis loop, it is determined that the vehicle is in the driving mode; If the required torque is output as 0 after a hysteresis loop, it is determined that the vehicle is in a braking mode.

4. The method according to claim 1, characterized in that: The step of obtaining the electric brake torque command and the air brake reference value according to the required torque and under the principle of feeding back as much electric energy as possible includes: Selecting the maximum value from the required torque and the maximum braking torque of the motor as the electric braking torque command; If the required torque exceeds the maximum braking torque of the motor, calculating a first torque difference between the required torque and the maximum braking torque; An air brake reference value is obtained according to the required torque, the first torque difference and the maximum braking power of the motor.

5. The method according to claim 4, characterized in that The step of obtaining the air brake reference value according to the required torque, the first torque difference and the maximum braking power of the motor includes: Calculating the difference between the required torque and the torque corresponding to the maximum braking power of the motor to obtain a second torque difference; The air brake reference value is calculated according to the smaller value of the first torque difference and the second torque difference, and the calculation relationship used is as follows: Among them, α airref is the air brake reference value, α cmd is the acceleration command, ΔT q is the torque difference, that is, the smaller value of the first torque difference and the second torque difference, T ref is the required torque.

6. The method according to claim 1, characterized in that The control parameter based on neural network adjustment adjusts the air brake reference value to obtain the air brake instruction, including: Obtain the difference between the acceleration command and the acceleration feedback value to obtain the acceleration difference; Taking the acceleration difference and the air brake compensation value output by the PI controller as input, the neural network is applied to adjust the parameters of the PI controller online with the goal of minimizing the mean square error of the acceleration difference; Obtaining an air brake compensation value by applying a PI controller according to the acceleration difference and the adjusted parameters; The air brake compensation value is applied to compensate the air brake reference value to obtain an air brake command.

7. The method according to claim 1, characterized in that The method further comprises: If it is determined that the vehicle is in the driving mode, the control directly outputs the required torque to the motor for motor driving.

8. A braking force distribution device for chassis control of an unmanned vehicle, characterized in that: The device comprises: A demand torque acquisition unit, used for obtaining the demand torque by applying a vehicle dynamics model and a PI controller according to an acceleration command; a mode determination unit, configured to determine whether the vehicle is in a driving mode or a braking mode according to the required torque; The braking force distribution unit is used to obtain the electric braking torque command and the air braking reference value according to the required torque and the principle of feeding back as much electric energy as possible if the vehicle is determined to be in braking mode, and adjust the air braking reference value based on the control parameters adjusted by the neural network to obtain the air braking command.

9. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein: When the processor executes the program, the method according to any one of claims 1 to 7 is implemented.

10. A computer storage medium, characterized in that: The storage medium stores at least one executable instruction, and the executable instruction enables the processor to execute the method according to any one of claims 1 to 7.

Images