A method, system and medium for deceleration closed-loop control of an intelligent driving vehicle

Abstract

The application relates to a deceleration closed-loop control method, system and medium of an intelligent driving vehicle, and the method comprises the following steps: U1. outputting a vehicle state according to required deceleration, calculating required driving or braking torque, and combining slope resistance, air resistance and rolling resistance to calculate data information of driving torque or braking torque required to be provided by a motor; U2. inputting the data information of the driving torque to a motor controller to output control current, calculating braking oil pressure according to the data information of the braking torque, outputting the braking oil pressure to an Ebooster to obtain actual oil pressure, forming a current closed loop according to real-time current measured by a current sensor, and forming an oil pressure closed loop according to real-time braking oil pressure measured by a braking pressure sensor. The application not only ensures that the automatic driving vehicle can travel more closely along a road track, but also improves the safety of vehicle driving and the comfort experience of passengers.

Description

Technical Field

[0001] This invention relates to the field of autonomous driving technology, and in particular to a method, system, and medium for closed-loop deceleration control of intelligent driving vehicles. Background Technology

[0002] Most current vehicle control systems focus on speed or distance control. However, for autonomous vehicles, which operate in complex road environments, braking control demands more. Excessive braking can affect passenger comfort, while insufficient braking can lead to collisions. Furthermore, the acceleration and deceleration range of vehicle control is limited. Acceleration control is necessary for starting, accelerating, stopping, trajectory tracking, speed following, and maintaining distance. Closed-loop control can better ensure that autonomous vehicles adhere to the road trajectory. Therefore, how to precisely control and adjust the braking of autonomous vehicles has become a pressing issue that needs to be addressed. Summary of the Invention

[0003] In view of the above problems, the present invention provides a deceleration closed-loop control method, system and medium for intelligent driving vehicles, which not only ensures that autonomous driving vehicles can drive more closely to the road trajectory, but also improves the safety of vehicle driving and the comfort experience of passengers.

[0004] To achieve the above and other related objectives, the present invention provides the following technical solution: A deceleration closed-loop control method for an intelligent driving vehicle, the method comprising: U1. Output the vehicle status based on the required deceleration, calculate the required drive or braking torque, and combine the data information of the drive torque or braking torque that the motor needs to provide, based on the slope resistance, air resistance and rolling resistance. U2. Input the driving torque data information to the motor controller to output control current, calculate the brake oil pressure based on the braking torque data information and output it to Ebooster to obtain the actual oil pressure, form a current closed loop based on the real-time current measured by the current sensor, and form an oil pressure closed loop based on the real-time brake oil pressure measured by the brake pressure sensor. U3. Based on the control current, the brake fluid pressure, the real-time current, and the real-time brake fluid pressure, a regression prediction algorithm based on the Flying Fox optimized BRANN model is used to predict the vehicle's brake fluid pressure, and the predicted vehicle brake fluid pressure data information is obtained. U4. Based on the predicted vehicle brake oil pressure data, calculate the braking torque and braking force of the front and rear wheels, and combine the slope resistance, air resistance and rolling resistance to obtain the actual deceleration, forming a deceleration closed loop.

[0005] Furthermore, the method also includes: U5. At the same time, a speed closed loop is formed based on the vehicle's initial speed and the actual speed obtained by the vehicle.

[0006] Furthermore, the distance is obtained based on the speed, the deceleration, and the time, forming a distance closed loop for the vehicle.

[0007] Furthermore, in step U3, the prediction of the vehicle's brake fluid pressure using a regression prediction algorithm based on the Flying Fox optimized BRANN model includes: U31. Initialization: Randomly generate the initial population for the Flying Fox algorithm, with each individual representing a set of parameters for the BRANN model; U32. Fitness Evaluation: Based on the prediction accuracy of the BRANN model for each individual, a fitness value function G is constructed. , Where x is the prediction accuracy of the BRANN model for each individual, and α, β and λ are weighting coefficients. The fitness values ​​of individuals in the population are calculated to obtain the fitness data of individuals in the population. U33. Position Update: According to the update strategy of the Flying Fox algorithm, the position of each individual is updated iteratively, that is, the parameters of the BRANN model are updated; U34. Termination condition judgment: If the maximum number of iterations is reached, the Flying Fox algorithm terminates; otherwise, return to step U32, output the BRANN model parameters and corresponding prediction accuracy corresponding to the optimal individual, and obtain the optimized BRANN model. U35. Based on the optimized BRANN model, the control current, the brake fluid pressure, the real-time current, and the real-time brake fluid pressure are input to predict the vehicle's brake fluid pressure, thereby obtaining the predicted vehicle brake fluid pressure data.

[0008] Furthermore, in step U32, the constraints on the weighting coefficients α, β, and λ are as follows: .

[0009] Furthermore, the update strategy of the Flying Fox algorithm is to sort the fitness values ​​of the individuals in the population based on the fitness value data information, and obtain the individuals with high fitness values. A high fitness value indicates high model prediction accuracy.

[0010] Furthermore, the BRANN model is an ensemble learning algorithm based on stochastic neural networks. It trains multiple stochastic neural networks with different structures and parameters, and then performs a weighted average of the prediction results of these networks to obtain the final prediction result. The control current, the brake oil pressure, the real-time current, and the real-time brake oil pressure are respectively input into stochastic neural networks with different structures and parameters to predict the brake oil pressure of the vehicle.

[0011] Furthermore, the BRANN model is an ensemble learning algorithm based on stochastic neural networks. It trains multiple stochastic neural networks with different structures and parameters, and then performs a weighted average of the prediction results of these networks to obtain the final prediction result. The control current, the brake oil pressure, the real-time current, and the real-time brake oil pressure are respectively input into stochastic neural networks with different structures and parameters to predict the brake oil pressure of the vehicle.

[0012] Furthermore, the control current input stochastic neural network obtains a predicted value R1, the brake oil pressure input stochastic neural network obtains a predicted value R2, the real-time current input stochastic neural network obtains a predicted value R3, and the real-time brake oil pressure input stochastic neural network obtains a predicted value R4, thus constructing a prediction function H. H = θR1 + ωR2 + σR3 + ρR4 θ, ω, σ, and ρ are weighting factors used to predict the vehicle's brake fluid pressure, thus obtaining the predicted brake fluid pressure data.

[0013] To achieve the above and other related objectives, the present invention also provides a deceleration closed-loop control system for an intelligent driving vehicle, including a computer device programmed or configured to perform the steps of any of the deceleration closed-loop control methods for intelligent driving vehicles.

[0014] To achieve the above and other related objectives, the present invention also provides a computer-readable storage medium storing a computer program programmed or configured to perform the deceleration closed-loop control method for any of the intelligent driving vehicles described herein.

[0015] The present invention has the following positive effects: This invention achieves closed-loop control of acceleration for intelligent driving vehicles, while simultaneously forming closed-loop control of speed, drive current, and brake oil pressure. This not only allows the vehicle to better follow the road trajectory but also makes the entire control process more precise and effective. It can directly control acceleration and form a control closed loop, which can more effectively ensure the vehicle's operating posture. Attached Figure Description

[0016] Figure 1This is a schematic diagram of the method flow of the present invention; Figure 2 This is a flowchart illustrating the regression prediction algorithm based on the Flying Fox optimized BRANN model of the present invention. Figure 3 This is a schematic diagram of the logic control framework of the present invention. Detailed Implementation

[0017] The exemplary embodiments of this disclosure are described below with reference to the accompanying drawings, including various details of the embodiments to aid understanding, and should be considered merely exemplary. Therefore, those skilled in the art will recognize that various changes and modifications can be made to the embodiments described herein without departing from the scope and spirit of this disclosure. Similarly, for clarity and brevity, descriptions of well-known functions and structures are omitted in the following description.

[0018] Example 1: As Figure 1 As shown, a deceleration closed-loop control method for an intelligent driving vehicle includes: U1. Output the vehicle status based on the required deceleration, calculate the required drive or braking torque, and combine the data information of the drive torque or braking torque that the motor needs to provide, based on the slope resistance, air resistance and rolling resistance. U2. Input the driving torque data information to the motor controller to output control current, calculate the brake oil pressure based on the braking torque data information and output it to Ebooster to obtain the actual oil pressure, form a current closed loop based on the real-time current measured by the current sensor, and form an oil pressure closed loop based on the real-time brake oil pressure measured by the brake pressure sensor. U3. Based on the control current, the brake fluid pressure, the real-time current, and the real-time brake fluid pressure, a regression prediction algorithm based on the Flying Fox optimized BRANN model is used to predict the vehicle's brake fluid pressure, and the predicted vehicle brake fluid pressure data information is obtained. U4. Based on the predicted vehicle brake oil pressure data, calculate the braking torque and braking force of the front and rear wheels, and combine the slope resistance, air resistance and rolling resistance to obtain the actual deceleration, forming a deceleration closed loop.

[0019] In this embodiment, the method further includes: U5. At the same time, a speed closed loop is formed based on the vehicle's initial speed and the actual speed obtained by the vehicle.

[0020] In this embodiment, the distance is obtained based on the speed, the deceleration, and the time, forming a distance closed loop for the vehicle.

[0021] In this embodiment, as Figure 2As shown, in step U3, the prediction of the vehicle's brake fluid pressure using a regression prediction algorithm based on the Flying Fox optimized BRANN model includes: U31. Initialization: Randomly generate the initial population for the Flying Fox algorithm, with each individual representing a set of parameters for the BRANN model; U32. Fitness Evaluation: Based on the prediction accuracy of the BRANN model for each individual, a fitness value function G is constructed. , Where x is the prediction accuracy of the BRANN model for each individual, and α, β and λ are weighting coefficients. The fitness values ​​of individuals in the population are calculated to obtain the fitness data of individuals in the population. U33. Position Update: According to the update strategy of the Flying Fox algorithm, the position of each individual is updated iteratively, that is, the parameters of the BRANN model are updated; U34. Termination condition judgment: If the maximum number of iterations is reached, the Flying Fox algorithm terminates; otherwise, return to step U32, output the BRANN model parameters and corresponding prediction accuracy corresponding to the optimal individual, and obtain the optimized BRANN model. U35. Based on the optimized BRANN model, the control current, the brake fluid pressure, the real-time current, and the real-time brake fluid pressure are input to predict the vehicle's brake fluid pressure, thereby obtaining the predicted vehicle brake fluid pressure data.

[0022] In this embodiment, in step U32, the constraints on the weighting coefficients α, β, and λ are as follows: .

[0023] In this embodiment, the update strategy of the Flying Fox algorithm is to sort the fitness values ​​of the individuals in the population according to the fitness value data information, and obtain the individuals with high fitness values. A high fitness value indicates high model prediction accuracy.

[0024] In this embodiment, the BRANN model is an ensemble learning algorithm based on stochastic neural networks. It trains multiple stochastic neural networks with different structures and parameters, and then performs a weighted average of the prediction results of these networks to obtain the final prediction result. The control current, the brake oil pressure, the real-time current, and the real-time brake oil pressure are respectively input to stochastic neural networks with different structures and parameters to predict the brake oil pressure of the vehicle.

[0025] In this embodiment, the BRANN model is an ensemble learning algorithm based on stochastic neural networks. It trains multiple stochastic neural networks with different structures and parameters, and then performs a weighted average of the prediction results of these networks to obtain the final prediction result. The control current, the brake oil pressure, the real-time current, and the real-time brake oil pressure are respectively input to stochastic neural networks with different structures and parameters to predict the brake oil pressure of the vehicle.

[0026] In this embodiment, the control current input random neural network obtains a predicted value R1, the brake oil pressure input random neural network obtains a predicted value R2, the real-time current input random neural network obtains a predicted value R3, and the real-time brake oil pressure input random neural network obtains a predicted value R4, thus constructing a prediction function H. H = θR1 + ωR2 + σR3 + ρR4 θ, ω, σ, and ρ are weighting factors used to predict the vehicle's brake fluid pressure, thus obtaining the predicted brake fluid pressure data.

[0027] Example 2: Based on the deceleration closed-loop control method for an intelligent driving vehicle in Example 1, the present invention will be further explained and described below.

[0028] like Figure 1 As shown, a deceleration closed-loop control method for an intelligent driving vehicle includes: U1. Output the vehicle status based on the required deceleration, calculate the required drive or braking torque, and combine the data information of the drive torque or braking torque that the motor needs to provide, based on the slope resistance, air resistance and rolling resistance. U2. Input the driving torque data information to the motor controller to output control current, calculate the brake oil pressure based on the braking torque data information and output it to Ebooster to obtain the actual oil pressure, form a current closed loop based on the real-time current measured by the current sensor, and form an oil pressure closed loop based on the real-time brake oil pressure measured by the brake pressure sensor. U3. Based on the control current, the brake fluid pressure, the real-time current, and the real-time brake fluid pressure, a regression prediction algorithm based on the Flying Fox optimized BRANN model is used to predict the vehicle's brake fluid pressure, and the predicted vehicle brake fluid pressure data information is obtained. U4. Based on the predicted vehicle brake oil pressure data, calculate the braking torque and braking force of the front and rear wheels, and combine the slope resistance, air resistance and rolling resistance to obtain the actual deceleration, forming a deceleration closed loop.

[0029] In this embodiment, as Figure 3As shown, S represents distance, V represents speed, a represents acceleration, P represents brake fluid pressure, des represents desired, act represents actual, Td represents torque of drive, Tb represents torque of brake, and Id represents current of drive or brake (considering energy recovery). The logical judgment is: if the difference = 0, maintain (vehicle speed remains unchanged); if the difference > 0, drive; if the difference < 0, brake. P_act & Id_act are measured by sensors.

[0030] In this embodiment, the present invention provides a deceleration closed-loop control system for an intelligent driving vehicle, including a computer device programmed or configured to perform the steps of any of the deceleration closed-loop control methods for intelligent driving vehicles.

[0031] In this embodiment, the present invention provides a computer-readable storage medium storing a computer program programmed or configured to perform the deceleration closed-loop control method for any of the intelligent driving vehicles described herein.

[0032] In this embodiment, any reference to memory, storage, database, or other media used in the embodiments provided in this application may include non-volatile and / or volatile memory. Non-volatile memory may include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM), or flash memory. Volatile memory may include random access memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in various forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), dual data rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), synchronous link DRAM (SLDRAM), RAMbus direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and RAMbus dynamic RAM (RDRAM), etc.

[0033] In summary, this invention not only ensures that autonomous vehicles can travel more closely to the road trajectory, but also improves vehicle driving safety and passenger comfort.

[0034] The specific embodiments described above do not constitute a limitation on the scope of protection of this disclosure. Those skilled in the art should understand that various modifications, combinations, sub-combinations, and substitutions can be made according to design requirements and other factors. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of this disclosure should be included within the scope of protection of this disclosure.

Claims

1. A closed-loop deceleration control method for an intelligent driving vehicle, characterized in that, The method includes: U1. Output the vehicle status based on the required deceleration, calculate the required drive or braking torque, and combine the data information of the drive torque or braking torque that the motor needs to provide, based on the slope resistance, air resistance and rolling resistance. U2. Input the driving torque data information to the motor controller to output control current, calculate the brake oil pressure based on the braking torque data information and output it to Ebooster to obtain the actual oil pressure, form a current closed loop based on the real-time current measured by the current sensor, and form an oil pressure closed loop based on the real-time brake oil pressure measured by the brake pressure sensor. U3. Based on the control current, the brake fluid pressure, the real-time current, and the real-time brake fluid pressure, a regression prediction algorithm based on the Flying Fox optimized BRANN model is used to predict the vehicle's brake fluid pressure, and the predicted vehicle brake fluid pressure data information is obtained. U4. Based on the predicted vehicle brake fluid pressure data, calculate the braking torque and braking force of the front and rear wheels, and combine the slope resistance, air resistance and rolling resistance to obtain the actual deceleration, forming a deceleration closed loop; In step U3, the prediction of the vehicle's brake fluid pressure using a regression prediction algorithm based on the Flying Fox optimized BRANN model includes: U31. Initialization: Randomly generate the initial population for the Flying Fox algorithm, with each individual representing a set of parameters for the BRANN model; U32. Fitness Evaluation: Based on the prediction accuracy of the BRANN model for each individual, a fitness value function G is constructed. ， Where x is the prediction accuracy of the BRANN model for each individual, and α, β and λ are weighting coefficients. The fitness values ​​of individuals in the population are calculated to obtain the fitness data of individuals in the population. U33. Position Update: According to the update strategy of the Flying Fox algorithm, the position of each individual is updated iteratively, that is, the parameters of the BRANN model are updated; U34. Termination condition judgment: If the maximum number of iterations is reached, the Flying Fox algorithm terminates; otherwise, return to step U32, output the BRANN model parameters and corresponding prediction accuracy corresponding to the optimal individual, and obtain the optimized BRANN model. U35. Based on the optimized BRANN model, the control current, the brake fluid pressure, the real-time current, and the real-time brake fluid pressure are input to predict the vehicle's brake fluid pressure and obtain the predicted vehicle brake fluid pressure data. In step U32, the constraints on the weighting coefficients α, β, and λ are as follows: 。 2. The deceleration closed-loop control method for intelligent driving vehicles according to claim 1, characterized in that, The method further includes: U5. At the same time, a speed closed loop is formed based on the vehicle's initial speed and the actual speed obtained by the vehicle.

3. The deceleration closed-loop control method for intelligent driving vehicles according to claim 2, characterized in that: The distance is obtained based on the speed, the deceleration, and the time, forming a distance closed loop for the vehicle.

4. The deceleration closed-loop control method for intelligent driving vehicles according to claim 1, characterized in that, The update strategy of the Flying Fox algorithm is to sort the fitness values ​​of the individuals in the population based on the fitness value data information, and obtain the individuals with high fitness values. A high fitness value indicates high model prediction accuracy.

5. The deceleration closed-loop control method for intelligent driving vehicles according to claim 1, characterized in that: The BRANN model is an ensemble learning algorithm based on stochastic neural networks. It trains multiple stochastic neural networks with different structures and parameters, and then performs a weighted average of the prediction results of these networks to obtain the final prediction result. The control current, the brake oil pressure, the real-time current, and the real-time brake oil pressure are respectively input into stochastic neural networks with different structures and parameters to predict the vehicle's brake oil pressure.

6. The deceleration closed-loop control method for intelligent driving vehicles according to claim 5, characterized in that: The control current input random neural network obtains a predicted value R1, the brake oil pressure input random neural network obtains a predicted value R2, the real-time current input random neural network obtains a predicted value R3, and the real-time brake oil pressure input random neural network obtains a predicted value R4. A prediction function H is then constructed. H = θR1 + ωR2 + σR3 + ρR4 θ, ω, σ, and ρ are weighting factors used to predict the vehicle's brake fluid pressure, thus obtaining the predicted brake fluid pressure data.

7. A deceleration closed-loop control system for an intelligent driving vehicle, comprising computer equipment, characterized in that, The computer device is programmed or configured to perform the steps of the deceleration closed-loop control method for an intelligent driving vehicle as described in any one of claims 1 to 6.

8. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a computer program that is programmed or configured to perform the deceleration closed-loop control method for an intelligent driving vehicle as described in any one of claims 1 to 6.

Images